Zool. Anz. 239 (2000): 147-178 © by Urban & Fischer Verlag ZOOLOGISCHER http://www.urbanfischer.de/journals/zoolanz ANZEIGER

The Mandible of Silverfish (Insecta: ) and (Ephemeroptera): Its Morphology and Phylogenetic Significance1

Arnold H. STANICZEK

Eberhard-Karls-Universität Tübingen, Zoologisches Institut, Lehrstuhl für Spezielle Zoologie, Tübingen, Germany

Abstract. The mandibles of several larvae have been investigated in order to reconstruct the groundplan of the mandibular muscles and of the mandibular articulation of Ephemeroptera. The results indicate the presence of three points of mandibular attachment to the cranium in the groundplan of Ephemeroptera. The posterior point of articulation in mayflies corresponds to the posterior (primary) mandibular joint of other . The remaining two points of attachment together form an anterior articulation complex made up of an anterolateral and a postero• medial part. These attachment points of the mayfly mandible are compared with the condition in silverfish. It is shown that in Zygentoma there is a similar anterior articulation complex present. This is interpreted as a groundplan character of Dicondylia. The posteromedial part of this complex in both Zygentoma and Ephemeroptera probably is the homologue of the anterior (secondary) mandibular joint of the remaining Dicondylia. The anterolateral part is secondarily modified in mayflies, where it forms a device to withstand the chewing pressure during adduction of the mandible. The primary mandibular articulation in both mayflies and silverfish is elongated and located dorsally to the secondary one, which results in an oblique axis of mandibular movement. In contrast to this plesiomorphic ar• rangement, the remaining Dicondylia have developed an almost horizontal axis of mandibular movement as well as a tight ball-and-socket type of mandibular articulation. As a consequence of the tightened mandibular articulation, the anterior articulation complex has been simplified in these groups, and a subgenal ridge is developed. The man• dibular musculature of Ephemeroptera and Zygentoma is plesiomorphic in retaining several mandibulo-tentorial muscles as well as two cranial abductor muscles. In all other pterygote lineages the dorsal tentorial muscles are lost, and only a single cranial abductor muscle is present. The mandibular muscles of the silverfish Tricholepidion gertschi show additional plesiomorphies, namely the lack of a mandibulo-hypopharyngeal muscle and the retention of a ligamentous transverse tendon. The groundplan of the dicondyle mandible is reconstructed, and the phy• logenetic significance of this character complex is discussed. The mandibles of both Zygentoma and Ephemeroptera do not show any traces of segmentation, thus the assumption of a telognathic insect mandible thereby cannot be supported. The distribution of all investigated mandibular characters support a sistergroup rela• tionship between Ephemeroptera and the remaining pterygote orders, the .

Key words. Dicondylia, phylogeny, phylogenetic systematics.

1. INTRODUCTION ent results and conclusions regarding the mandibular articulation. The subimagines and imagines of mayflies do not con• SNODGRASS (1950, 1951) recognized just one mandi• tinue feeding, and their mandibles as well as their other bular joint in the mayfly mandible, whereas other work• mouthparts are highly atrophied in these life stages. ers notified two mandibular articulations (BÖRNER However, the aquatic larvae still bear well-developed 1909; KUKALOVÁ-PECK 1985, 1991; SCHÖNMANN 1981) mandibles. Several authors have investigated the head in different species. Some authors even stated three morphology of mayfly larvae, but came to quite differ• separate articulations (ARENS 1989; BROWN 1961; STRENGER 1953, 1970, 1975, 1977, 1979). 1 This paper is dedicated to Professor Dr. Karl-Ernst Lauter- Similarily, descriptions of silverfish mandibles vary bach (Bielefeld, Germany). considerably. While all authors agree on the presence of

0044-5231/00/239/02-147 $ 12.00/0 148 A. H. STANICZEK

an anterior articulation, SNODGRASS (1950, 1952, 1960) The larvae of the New Zealand genus Nesameletus mainly repeatedly referred to the anterior condyle as being lo• feed on diatoms (own observation). cated at the gena, whereas CHAUDONNERET (1950) ob• served an anterior condyle on the laterodorsal angle of • Ameletopsis perscitus Eaton, 1899 (Ameletopsidae) the clypeus. The New Zealand species of the amphinotic distributed In recent publications some authors also claim the Ameletopsidae belongs to one of the very few mayfly fami• presence of a (at least vestigial) segmentation of the lies with larvae with exclusively carnivorous habits. Amele- topsids are known to engulf other water (CAMPBELL mandible in both Zygentoma (KRAUS 1998; KUKALOVÁ- 1985). The gut contents of the larger specimens used for this PECK 1998) and Ephemeroptera (KUKALOVÁ-PECK study consisted mainly larvae of Nesameletus and 1998). Interpreting the mandible as composed of the larvae. entire telopodite, they revived the dispute between A. perscitus was chosen for this study to observe the impact of MANTON's (1964) idea of a 'whole-limb jaw' in the a totally different nutrition on the anatomy of the mandible. groundplan of Tracheata and the concept of a gnathoba- sic mandible in these groups (LAUTERBACH 1972a). These obvious inconsistencies in both description and 2.2. Additionally used material for interpretation of the mandible in mayflies and silverfish comparisons led to the present study. It is aimed to reconstruct the condition of the mandible in the respective ground- Zygentoma: plans, and to present new arguments for the still contro• Nicoletiidae: Cubacubana spec. versially discussed early phylogenetic branching events Lepismatidae: Lepisma saccharina Linné, 1758 of the Dicondylia. Ctenolepisma spec.

Ephemeroptera: 2. MATERIALS AND METHODS Ameletidae: Ameletus inopinatus Eaton, 1887 Metrehtus balcanicus (Ulmer, 1920) 2.1. Specimens examined Ametropodidae: Ametropus fragilis Albarda, 1878 Baetidae: Baetis rhodani (Pictet, 1843) Zygentoma Baetis vernus Curtis, 1834 • Tricholepidion gertschi Wygodzinsky, 1961 Centroptilum luteolum (Müller, 1776) (Lepidothrichidae) Coloburiscidae: Coloburiscus humeralis (Walker, 1853) Ephemerellidae: Serratella ignita (Poda, 1761) This North American relic silverfish is known to possess Torleya major (Klapalek, 1905) some unique plesiomorphic traits not only within Zygentoma Heptageniidae: Ecdyonurus venosus (Fabricius, 1775) (WYGODZINSKY 1961), but also regarding the groundplan of Leptophlebiidae: Habroleptoides confusa Sartori & Jacob, Dicondylia (KRISTENSEN 1998). For instance it is the sole 1986 known recent species of all dicondyle insects that has retained Metretopodidae: Metretopus borealis (Eaton, 1901) the mandibular transverse ligament, otherwise well known Oligoneuriidae: Oligoneuriella rhenana (Imhoff, 1852) from Archaeognaths and entognath insect orders (BOUDRE- Oniscigastridae: Oniscigaster distans Eaton, 1886 AUX 1979a). However, its entire mandibular musculature as Polymitarcyidae: Ephoron virgo (Olivier, 1791) well as its mandibular articulations have not previously been Potamanthidae: Potamanthus luteus (Linne, 1767) subject to a close examination. Rallidentidae: Rallidens mcfarlanei Penniket, 1966 Siphlaenigmatidae: Siphlaenigma janae Penniket, 1962 Ephemeroptera Siphlonuridae: Siphlonurus croaticus Ulmer, 1920 Siphlonurus lacustris (Eaton, 1870) • Oniscigaster wakefieldi McLachlan, 1873 (Oniscigastridae) This species represents one of the two New Zealand represen• tatives of a mayfly family with amphinotic distribution. Its : imago is considered to resemble the mayfly groundplan in Aeshnidae: Aeshna cyanea Müller, 1764 many aspects (MCLACHLAN 1873; PENNIKET 1962; KLUGE et Calopterygidae: Calopteryx virgo Linne, 1758 al. 1995), so it seemed worthwhile to have also a closer look at its larva. The larva feeds mainly on detritus (MCLEAN 1970). :

• Nesameletus sp. (Nesameletidae) Taeniopterygidae: Taeniopteryx spec. Nesameletidae is another mayfly taxon with amphinotic dis• tribution. Its phylogenetic relationship within Ephemeroptera Ensifera: is up to now not satisfactory resolved (KLUGE et al. 1995). Gryllidae: Acheta domesticus Linné, 1767 The Mandible of Silverfish and Mayflies 149

2.3. Methods M16 dorsal suspensor muscle of stomodaeum (M. vertico- pharyngealis) Fixation. The material used was preferably fixed by a mixture M17 lateral dilatator muscle of pharynx (Dilatator pharyn- of 95% ethanol, 35% formaldehyde, and acetic acid (66:33:10). gis lateralis primus) After 24 hours or longer it was transferred to 80% ethanol. Some M18 suspensor muscle of tentorium species (Ametropus, Metretopus) were fixed in ethanol only. M19 antennal muscles (Mm. antennarum) Manual dissection. Larvae were dissected under 80% etha• M20 tentorial adductor muscle of cardo (M. tentoriocardi- nol on a layer of paraffin in a Petri dish. To observe muscles, nalis) they were stained with basic fuchsin. To observe cuticular M21 tentorial adductor muscle of stipes (M. tentoriostipi- structures, specimens were kept in 10% potassium hydroxide talis) under room temperature for several days, until the soft tissues M22 cranial adductor muscle of galeolacinia (M. cranio- dissolved. Then the cuticle was stained with Chlorazol Black. lacinialis) To duplicate the mandible movements, the mandible as well M23 tentorial connective muscle as its muscles were manipulated by a Dumont forceps. aac anterior articulation complex Light microscopy. Specimens used for microscopic section• ab antennal base ing were dehydrated in ethanol and then stored three times at ac anterior condylus 50°C in propan-2-ol for 24 hours each time. Then the material aclp anteclypeus was gradually transferred to paraffin at 50 °C and finally antenna transferred to Paraplast Plus™ at 60 °C. There the specimens alp anterolateral part of anterior articulation complex were kept under vacuum conditions for 24 hours to optimize ao aorta their penetration. Finally the material was embedded in Para• ap apodeme plast Plus™. Sections of 5-7 µm thickness were obtained ata anterior tentorial arm with a rotation microtome. Sections were stained with De- atp anterior tentorial pit lafield's hematoxylin, counter-stained with erythrosin, and observed with a Zeiss-Axioplan microscope. Photographs br supraoesophageal ganglion were taken with a Zeiss-MC 100 camera. cc corpora cardiaca Scanning electron microscopy (SEM). Specimens were de• clp clypeus hydrated through a stepwise immersion in ethanol and ace• conn circumoesophaeal connective tone, and then dried by critical point drying. The mounted ct cuticula material was coated with a 20 nm Au/Pd layer and examined cte corpotentorium with a Cambridge Stereoscan 250 MK 2 scanning electron dta dorsal tentorial arm microscope at 10 kV. epd epidermis eph epipharynx fb fat body 2.4. Abbreviations fr frons ge gena

Ml cranial adductor muscle of mandible (M. cranio- gei inflected part of gena mandibularis internus) hyp hypopharynx M2 anterior cranial abductor muscle of mandible (M. inc incisivi cranio-mandibularis externus anterior) lb labium M3 posterior cranial abductor muscle of mandible (M. lbr labrum cranio-mandibularis externus posterior) le lateral compound eye M4 dorsolateral tentorial muscle (M. tentorio-mandibu- If lateral fold laris externus dorsalis) lig transverse ligament M5 ventrolateral tentorial muscle (M. tentorio-mandibu- lo lobus opticus laris externus ventralis) md mandible M6 dorsomedial tentorial muscle (M. tentorio-mandibu- mg mandibular groove laris internus dorsalis) mil mandibular lateral ledge M7 ventromedial tentorial muscle (M. tentorio-mandibu- mmo morphological mouth opening laris internus ventralis) mn mandibular notch M8 hypopharyngeal protractor muscle (M. mandibulo-hy- mo mola popharyngealis) mx maxilla M9 adductor muscle of labrum (M. frontolabralis) Nant antennal nerve M10 abductor muscle of labrum (M. frontoepipharyngealis) Nlbr labral nerve Ml 1 dilatator muscle of cibarium (M. clypeopalatalis) oc occiput M12 mouth retractor muscle (Retractor angulorum oris) ol lateral ocellus M13 dorsal dilatator muscles of pharynx (Dilatator pharyn- om median ocellus gis) pa posterior articulation M14 circular pharynx muscle (M. anularis stomodaei) pc posterior condylus M15 longitudinal pharynx muscles (M. longitudinalis sto• pclp postclypeus modaei) pg postgena 150 A. H. STANICZEK plb palpus labialis ptp posterior tentorial pit pmx palpus maxillaris sc scapus poc postocciput scr sclerotised cranial ridge of alp pocph postoccipital phragma sg subgena pocrd postoccipital ridge sgrd subgenal ridge pp processus paratentorialis of anterior articulation com- smr sclerotised mandibular ridge of pmp plex stom stomodaeum pmp posteromedial part of anterior articulation complex tp tentorial plate pro prostheca tr trachea pta posterior tentorial arm vx vertex

Fig. 1. O. wakefieldi, head of larva, lateral view, SEM.

Fig. 2. O.wakefieldi, left mandible, poste• rior view, SEM. Arrows indicate desclero- tised area.

Fig. 3. O. wakefieldi, left mandible, lateral view, SEM. The Mandible of Silverfish and Mayflies 151

3. RESULTS 3.1. Oniscigaster wakefieldi McLachlan, 1873

Mandible The larva of O. wakefieldi bears orthognathous mouth- parts (Fig. 1). (The term 'orthognathous' rather than 'hypognathous' is used herein, because the term 'hy- pognathous' is also used as a synonym for the term 'opisthognathous' by various authors (see VON KÉLER 1963 but NICHOLS et al. 1989)). The mandible is well developed and heavily sclerotized. It resembles the mandible of an orthopteroid insect, but the mandible's outer basal edge declines from its caudal to its frontal end. Each mandible bears two well-developed incisivi, a prostheca and a molar region (Fig. 2). The mandibles are sharply bent inwards frontally, so that their ante• rior surfaces lie behind the epipharynx in the trans• verse plane. Externally only the posterior mandibular articulation is partly visible (Fig. 1). It is located at the hind angle of the mandible (Figs. 1, 3, 5). The postero• lateral part of the mandible is basally elevated forming a laterally compressed and elongated roller-like condyle (Figs. 3-5). The roller has an oblique position with an axis that lies at an angle of about 55° to the transverse plane. The medial end of the roller is lo• cated anterior and also slightly dorsal to its lateral end. The central part of the roller runs inside a socket that is formed by the postgena. The socket is shaped as a hol• low groove that is bordered medially as well as later• ally by two regions that are heavily sclerotized (Fig. 5). Fig. 4. O. wakefieldi, mazerated head from anterior. The At the basal frontolateral edge of each mandible an an• head capsule has been partly removed to enable a view onto terior articulation complex (aac) is located: its antero• tentorium, inflected gena, and anterolateral part of the ante• lateral part (alp) is represented by an elongated mandi• rior articulation complex. bular notch into which the inner margin of the inflected Fig. 5. O. wakefieldi, mazerated head from posterior, view onto tentorium and right mandible. Other mouthparts have cranium projects (Figs. 4, 6). This mandibular notch been removed to enable a look onto the cranial part of the an• (mn) is formed as an oblique, elongated impression of terior mandibular articulation and the anterior tentorial pit. the mandibular body, that is situated distally to the basal mandibular edge, and thus well separated from the latter (Figs. 3, 7). Hereby the anterolateral border of the notch (mll), which is heavily sclerotized, partly overlaps the notch itself, thus creating a hollow groove (Figs. 3, 7). The sclerotized cranial ridge (scr, Fig. 5) at the margin of the inflected gena (Figs. 32c, d) perfectly fits into this groove, together forming a loose hinge joint (Fig 6). This is the structure that has been de• scribed in different mayfly larvae as anterior articula• tion, slider, or guiding structure by previous authors (STRENGER 1953-1979; BROWN 1961; KUKALOVÁ-PECK 1985, 1991; ARENS 1989). Immediately behind and dorsally to this joint the pos• teromedial part (pmp) of the anterior articulation com• Fig. 6. O. wakefieldi, left mandible in lateral view. The lat• plex is located. The basal outer rim of the mandible is eral head capsule has been partly removed to enable a look heavily sclerotized and slightly indented in this region, onto the anterior articulation complex. 152 A. H. STANICZEK forming a small, saddle-like socket (Fig. 7, smr). Dor- Tentorium sally, the cranial processus paratentorialis (pp, see also below) is attached to this part of the mandibular rim. The anterior tentorial pits are located at the inner edge forming another point of articulation. In this way the of the inflected ventral parts of the genae, and thus not mandibular rim forms an acute angle with the roller• visible externally. Only after removing all the mouth- like processus paratentorialis (Fig. 6). parts a view from posterior reveals the position of the The axis of mandibular movement runs between the anterior tentorial pits (Figs. 5, 8, atp). Their position is posterior articulation and the posteromedial part of the confluent with the ventrolateral border of the sclero• anterior articulation complex. The posterior articulation tized cranium. The anterior tentorial pits mark the point lies dorsally and laterally of the posteromedial part of of invagination of the anterior tentorial arms (ata) that the anterior articulation complex, thus creating an are well developed (Figs. 4. 5). Close to the anterior oblique axis of mandibular movement. edge of the anterior tentorial pit an elongated, heavily

Fig. 7. O. wakefieldi, right mandible, anterior articulation complex, latero- dorsal view, SEM. Arrows indicate the membraneous connection of the man• dible with the basal cranial border. The progression of the membrane thus marks the border between external and internal environment.

Fig. 8. O. wakefieldi, cranial part of left anterior articulation complex, ante• rior tentorial pit, and processus para• tentorialis in posterior view, SEM. Ar• rows indicate the membraneous con• nection of the cranium with the basal mandibular rim. The progression of the membrane thus marks the border be• tween external and internal environ• ment. The Mandible of Silverfish and Mayflies 153

sclerotized processus is present (Figs. 5, 6, 8, 32f). This different portions of M1. Both sides of the major plate structure is termed herein as processus paratentorialis are points of attachment for those three portions, that (pp), as it is not formed by the invaginated anterior ten• extend to the vertex (Fig. 32a). Other portions that in• torial arm itself, but it is clearly an external structure sert at both sides of the minor tendinous plate, extend to formed by the cranium and positioned laterally of the the postoccipital ridge as well as to the occiput and pos- craniomandibular basal connective membrane (Fig. 8). tocciput itself (Figs. 9-10). All portions of this muscle This processus paratentorialis is part of the anterior ar• serve as adductor of the mandible. ticulation complex, and articulates with the basal outer rim of the mandible. Its position near the anterior edge (M2) M. cranio-mandibularis externus anterior (ante• of the anterior tentorial pit is defined herein as antero- rior cranial abductor): condylate. (crm2) muscle cranio-mandibulaire 2 BITSCH 1963 The anterior tentorial arms arise well separated from (adcl) muscle adducteur cranien lateral CHAUDONNERET 1950 the clypeus in a somewhat posterior position and taper (r.m.a.) Remotor mandibulae anticus BÖRNER 1909 as they converge and become united with the posterior (3a) cranial abductor muscle BROWN 1961 tentorial arms (Fig. 4). The anterior tentorial arm is u- (25a) cranialer Mandibelabduktor SCHÖNMANN 1981 shaped in cross section (Figs. 29b, 32f-g). The anterior tentorial arm is also slightly twisted as it converges me• This relatively small and short abductor muscle inserts dially, where it splits off the dorsal arm (dta, Figs. 4, 5, at a short and tenuous apodeme immediately posterior 32a-f), that is formed as a broad plate-like extension. to the posteromedial part of the anterior articulation Immediately medial to the antenna the dorsal tentorial complex. It extends in an oblique direction dorsally and arm is attached to the epidermis by tonofibrillae. Be• posterior to its lateral origin at the gena just below the sides the extrinsic muscles of the antenna, some ventral compound eye (Figs. 9, 32g). This muscle serves to• mandibular muscles are also attached to it. gether with the following one as an abductor of the mandible. Additionally, its action probably tightens the attachment between the processus paratentorialis and Mandibular muscles its corresponding mandibular socket by compressing The mandibular muscles basically consist of two the two components of the joint. groups. The dorsal muscles take their origin on the cra• nium and insert by apodemes at the basal mandibular edge. The ventral muscles originate on the the tento• rium or hypopharynx and insert at the mandibular edge or at the body of the mandible.

For details on the homology of muscles see discussion. a) Dorsal muscles (Ml) M. cranio-mandibularis internus (cranial adductor muscle of mandible): (crml) muscle cranio-mandibulaire 1 BITSCH 1963 (adcm) muscle adducteur cranien médial + (abcmd) muscle abducteur cranien medial CHAUDONNERET 1950 (a.m.p.) Adductor mandibulae posticus BÖRNER 1909 (6) Dorsal adductor muscles BROWN 1961 (MA) dorsaler Mandibeladduktor, cranialer Mandibeladduk- tor STRENGER 1953, 1970, 1975, 1977 (MAa-MAd) dorsaler Mandibeladduktor SCHÖNMANN 1981

The cranial adductor muscle is the most powerful man• dibular muscle. It consists of several bundles of muscle fibres. Its apodeme rises as a strong tendon at the inner side of the basal mandibular rim, just between the molar region and the posterior articulation (Figs. 5, 11). This tendon soon splits off a second branch, that is Fig. 9. O. wakefieldi, lateral view of the larval head. The lat• somewhat thinner than the main tendon. Each branch is eral cranium has been partly removed to show the mandibular apically broadened forming a plate-like attachment for articulation and the lateral mandibular muscles. 154 A. H. STANICZEK

Fig. 11. O. wakefieldi, left mandible with musculature in posterior view. M5 is not visible, M8 has not been drawn.

Fig. 10. O. wakefieldi, lateral view of the larval head. The lateral cranium, parts of the lateral mandibular wall, and the muscles M2-M4 have been removed to enable a look onto the mandibular muscles M5-M8. (4) tentorial abductor muscles BROWN 1961 [in partim] (25') tentorialer Mandibelabduktor, Abductor tentoricus STRENGER 1953, 1970, 1975, 1977 (25') tentorialer Mandibelabduktor SCHÖNMANN 1981 (M3) M. cranio-mandibularis externus posterior (poste• This muscle consists of two adjacent portions. Both in• rior cranial abductor): sert on the mandibular body just below the tendon of (crm3) muscle cranio-mandibulaire 3 BITSCH 1963 (3), approximately half way between processus para• (abca) muscle abducteur cranien anterieur CHAUDONNERET tentorialis and posterior articulation (Fig. 9). Its upper 1950 portion originates on the dorsal tentorial arm, its lower (r.m.p.) Remotor mandibulae posticus BÖRNER 1909 portion on the anterior tentorial arm. This muscle prob• (3b) cranial abductor muscle BROWN 1961 ably serves as an abductor of the mandible. (25) cranialer Mandibelabduktor STRENGER 1953, 1970, 1975, 1977 (M5) M. tentorio-mandibularis externus ventralis (ven• (25b) cranialer Mandibelabduktor SCHÖNMANN 1981 trolateral tentorial muscle):

It inserts at the basal outer mandibular rim on a moder• (adml) BITSCH 1963 ately sized, plate-like apodeme that is located approxi• (adt) muscle adducteur tentorial principal CHAUDONNERET mately in the middle between the posteromedial part of 1950 the anterior articulation complex and the posterior articu• (a.m.t.) Adductor mandibulae tentoricus BÖRNER 1909 lation. It extends medially of the M. craniomandibularis (5) tentorial adductor muscles BROWN 1961 [in partim] externus anterior in an oblique direction to its broad ori• (26b) tentorialer Mandibeladduktor, Adductor tentoricus gin at the lateral part of the postoccipital ridge (Fig. 9) STRENGER 1970 This muscle serves as an abductor of the mandible. (26) tentorialer Mandibeladduktor, Adductor tentoricus STRENGER 1977 (26a) tentorialer Mandibeladduktor SCHÖNMANN 1981 b) Ventral muscles: This rather small muscle takes its origin at the anterior (M4) M. tentorio-mandibularis externus dorsalis (dor• tentorial arm close to the processus paratentorialis solateral tentorial muscle) (Figs. 9, 10). It extends in an oblique direction ventro- (ttm2) BITSCH 1963 caudad to its insertion on the distal area of the mandi• (abtl) muscle abducteur tentorial latéral CHAUDONNERET 1950 bular body. It assists in adducting the mandible. The Mandible of Silverfish and Mayflies 155

(M6) M. tentorio-mandibularis internus dorsalis (dor- 3.2. Nesameletus sp. somedial tentorial muscle): The mouthpart are slightly opisthognathous (Fig. 12). (ttm3) BITSCH 1963 (abtp) muscle abducteur tentorial postérieur CHAUDONNERET The pars incisiva is modified: instead of two incisivi 1950 there is a large gouge developed. Basally of the gouge (4) tentorial abductor muscles BROWN 1961 [in partim] the bipartite prostheca is present, but its lower part is (26') Adductor tentoricus STRENGER 1970. 1977 strongly reduced to a single bristle (Fig. 14). The pars (26c) tentorialer Mandibeladduktor SCHÖNMANN 1981 molaris is very similar to the condition in O. wakefieldi bearing several cuticular rows on its surface. Externally This muscle originates at the dorsal as well as the ante• only the posterior mandibular articulation is partly visi• rior tentorial arm below the areas of origin of (4) and ble, whose mandibular part is formed as a roller-like (7). It extends backwards to the upper inner end of the condyle. The cranial socket is accordingly shaped as a mandible, inserting at its inner rim just below the poste• hollow groove that is bordered medially as well as lat• rior mandibular condyle (Figs. 9, 10). It probably erally by two regions that are heavily sclerotized. serves as an adductor as well as a rotator of the man• At the basal frontolateral edge of each mandible the an• dible. terior articulation complex is located, that is repre• sented by an anterior and posterior part. Both parts are not exposed, but are easily visible from outside, as due (M7) M. tentorio-mandibularis internus ventralis (ven• to their heavy degree of sclerotization the dark colour• tromedial tentorial muscle): ing shines through the overlying lateral body wall. Its anterior part also features an elongated mandibular de• (ttml) muscle tentorio-mandibulaire 1 BITSCH 1963 (adt') muscle adducteur tentorial accessoire CHAUDONNERET pression (Fig. 14) into which the inner margin of the in• 1950 flected cranium fits. This hinge joint is somewhat (a.m.t.) Adductor mandibulae tentoricus BÖRNER 1909 [in shorter than the one in O.wakefieldi, but otherwise partim] there are no differences to observe. (5) tentorial adductor muscles BROWN 1961 [in partim] In contrast to this, the posterior part of the anterior ar• (26) tentorialer Mandibeladduktor. Adductor tentoricus ticulation complex in Nesameletus sp. is somewhat dif• STRENGER 1970 ferent in both shape and position. It is very well sepa• (26a) tentorialer Mandibeladduktor. Adductor tentoricus rated from the anterior part and located approximately STRENGER 1977 halfway between the latter and the posterior (primary) (26b) tentorialer Mandibeladduktor SCHÖNMANN 1981 mandibular articulation (Figs. 13. 14) at the posterior edge of the anterior tentorial pit. This condition is This rather big muscle originates below (4) on the ante• herein referred to as posterocondylate. rior tentorial arm (Fig. 10). It runs caudad and inserts The basal outer rim of the mandible in this region is on the medial wall of the mandibular cavity, just be• heavily sclerotized and elevated, forming an elongated tween the insertions of (1) and (6). It acts as an adductor pedestal with a plane dorsal surface (Fig. 14). Dorsally of the mandible. the likewise plate-like processus paratentorialis makes contact with the elevated mandibular rim, forming a broad and plane area of contact that is heavily sclero• (M8) M. mandibulo-hypopharyngealis (protractor mus• tized. The anterior tentorial arms are very similar to the cle of hypopharynx): ones in O.wakefieldi. They also exhibit an u-shaped (adm2) BITSCH 1963 ? cross-section. The position of the anterior tentorial pits (adh) muscle adducteur hypopharyngien CHAUDONNERET is also confluent with the ventrolateral border of the 1950 sclerotized cranium. (zgm.) Zungenmuskel BÖRNER 1909 (7) hypopharyngeal-mandibular muscle BROWN 1961 (29) STRENGER 1953. 1970. 1975 (29) SCHÖNMANN 1981 Mandibular muscles (Figs. 12-13)

The smallest muscle of the mandible takes its origin a) Dorsal muscles below the anterior articulation on the mandibular body (M1) M. cranio-mandibularis internus: (Figs. 9-10, 32g). It extends mediad and caudad and in• serts with a long thin tendon on the basal lateral edge of The cranial adductor muscle is also the biggest mandi• the hypopharynx. Though often described as an adduc• bular muscle. Its apodeme is elongated to serve as at• tor of the mandible, it could as well serve as protractor tachment for several muscle bundles that are all placed of the hypopharynx. in line and run parallel to each other from their origins 156 A. H. STANICZEK

Fig. 12. Nesameletus sp., lateral view of the larval head. The Fig. 13. Nesameletus sp., lateral view of the larval head. The lateral cranium has been partly removed to show the lateral cranium, parts of the lateral mandibular wall, and the mandibular articulation and the lateral mandibular muscles. muscles M2-M4 have been removed to enable a look onto the mandibular muscles M5-M8.

at the vertex. A smaller portion takes its origin at the (M5) M. tentorio-mandibularis externus ventralis: lateral postoccipital ridge. (M5) is somewhat bigger than the corresponding mus• cle in Oniscigaster. In addition its insertion is some• (M2) M. cranio-mandibularis externus anterior: what dorsal compared with the latter. Otherwise there The anterior cranial abductor muscle is very small, but are no big differences to observe. otherwise shows no difference from the respective mus• cle in O. wakefieldi. (M6)-(M8) are identical in shape and orientation to the respective ones in O. wakefieldi. (M3) M. cranio-mandibularis externus posterior: The posterior abductor muscle is somewhat shorter 3.3. Ameletopsis perscitus (Eaton, 1899) compared with the respective one in O. wakefieldi, but has basically the same attachment points and orienta• The head capsule of the carnivorous A. perscitus is tion. much broader than its thorax (Fig. 15). The entire head is dorsoventrally flattened and is capable of spacious movements in dorsoventral direction. b) Ventral muscles

(M4) M. tentorio-mandibularis externus dorsalis: Mandible This muscle is very similar in both shape and orienta• Basally, the mandible of A. perscitus is almost circular tion to the condition in O. wakefieldi and there are two in cross section. It is slightly curved inwards at its ante• different layers present. rior third (Fig. 16). The mola is absent, the lateral inci- The Mandible of Silverfish and Mayflies 157

Fig. 14. Nesameletus sp., left mandible in anterior view, SEM.

Fig. 15. A. perscitus, larval head in lateral view, SEM.

Fig. 16. A. perscitus. left mandible. The arrows indicate desclerotised strips at the base of incisivi, SEM. 158 A. H. STANICZEK

sivus (inc1) is extended to a dagger-like cusp with three anterior articulation complex is easily seen externally, prongs at its tip. The medial incisivus (inc2) is some• whereas the posterior mandibular articulation is cov• what shorter and lies well separated from the former ered by a duplicature of the ventrocaudal corner of the parallel to a thickened and massive prostheca, that has lateral cranium (Fig. 17). just few stout bristles scattered at its rounded tip. The At its posterior tip the mandible tapers into a tongue• prostheca as well as the medial incisivus point medially like condyle that is slightly curved. The condyle ex• to the mouth opening. Both incisivi are basally descle- tends almost in a parasagittal direction. Medially and rotised (Fig. 16, arrows) and thus to a certain extent laterally, it is embedded in a cranial socket shaped as a flexible. It takes no great effort to move the incisivi in• furrow. The medial border of the cranial socket is wards by manipulation. formed by the postocciput. The anterior part of the anterior articulation complex is flattened and barely visible as a flattened triangle. The Tentorium mandibular hollow groove is missing, only a shallow impression of the cuticula indicates the remnants of a The anterior tentorial arm is very thin and clasp-like, mandibular notch (Fig. 16). but it resembles the characteristic u-shaped cross cut. The posterior part of the anterior articulation is widely The processus paratentorialis is formed as a small separated from its anterior part and thus shows the pos- knob, that is attached medially to the respective man• terocondylate condition. It probably serves as the func• dibular socket. Medially the anterior tentorial arm tional anterior articulation. The basal mandibular rim is broadens widely to form a large and flat dorsal tento• strongly thickened and forms an extended sclerotised rial arm (Fig. 18). Dorsocaudal to the antennae the strip. With the mandible completely adduced, the pro• dorsal tentorial arm is attached to the cranium. A cor- cessus paratentorialis clings to the strip's caudal end porotentorium is not present: both anterior tentorial from medially. arms extend separated from each other as two almost The lateral cranial border does not overlap the basal invisible tiny chitinuous lamellae to the posterior ten• mandibular rim at this area, so the posterior part of the torium.

Fig. 17. A. perscitus, lateral view of the larval head. The lat• Fig. 18. A. perscitus, lateral view of the larval head. The lat• eral cranium has been partly removed to show the mandibular eral cranium, parts of the lateral mandibular wall, and the articulation and the lateral mandibular muscles. muscles M2-M3 have been removed to enable a look onto the medial mandibular muscle M7. The Mandible of Silverfish and Mayflies 159

Mandibular muscles (Figs. 17-19) ridge across the entire occiput. It extends ventrocraniad and crosses in its distal third M2 laterally. It inserts (Ml) M. cranio-mandibularis internus: without apodeme anterior to the insertion of M2 and immediately behind the posterior part of the anterior ar• The massive cranial adductor muscle inserts at a very ticulation complex at the lateral basal rim of the man• short apodeme, that invaginates at the frontomedial dible. basal rim of the mandible. It consists of just two large parts that originate at the ipsilateral vertex and at the (M4), (M5), (M6), (M8) are absent. postoccipital rim. (M7) M. tentorio-mandibularis internus ventralis: (M2) M. cranio-mandibularis externus anterior: The only ventral muscle of the mandible surrounds the The anterior cranial abductor muscle is greatly enlarged anterior tentorial arm from ventrally. Its origin lies on compared to the respective muscles of the other investi• both sides of the dorsal tentorial arm.The muscle ex• gated species. Also its origin as well as its insertion is tends from its origin in an oblique direction laterocau- somewhat different: dal and inserts at the hind third of the mandibular The anteroventral part of the muscle originates laterally body's medial side. of M19 at the upper side of the broadened dorsal tento• rial arm. In some of the investigated specimens this part of the muscle even overlaps the lateral border of the 3.4. Tricholepidion gertschi dorsal tentorial arm and extends to its lower side. The posterodorsal part of the muscle originates at the frons, closely behind the dorsal tentorial arm and fron- Mandible tolaterally to the lateral ocellus. The mandible of T. gertschi is elongated (Fig. 20). It Both parts of the muscle extend laterocaudally. Shortly bears a well developed pars incisiva and pars molaris in front of their common insertion, both parts of M2 (Fig. 27). The mandibular body has an approximately cross M3 medially. In contrast to all other examined circular cross-section, and the opening of the mandi• species, the point of insertion lies well separated from bular cavity is directed dorsomedially. From its poste• the posterior part of the anterior articulation process in rior to its anterior end the basal mandibular rim slopes the posterior third of the mandible, approximately in inwards and slightly downwards. The posterior (pri• between the processus paratentorialis and the posterior mary) articulation is located at the caudal basal end of articulation. An apodeme is not present. the mandibular rim (Fig. 26). Its condyle is formed as an elongated and heavily sclerotized ledge. It fits into (M3) M. cranio-mandibularis externus posterior: the slightly curved lower hind angle of the head, thus creating a ginglymus. At the basal frontolateral edge of The posterior cranial abductor muscle is a very big, flat the mandible an anterior articulation complex is located muscle, whose origin extends parallel to the occipital (Figs. 21, 24-25): Its lateral part comprises an elongated cuticular ledge of the mandible (mll) that fits into a kink of the ventrolat• eral margin of the clypeus. This mandibular lateral ledge is well separated from the basal mandibular edge. Medial to this sclerotised ledge the clypeal margin rests on the horizontally flattened mandibular body, that is slightly impressed in that region to form a shallow groove (mg), but without particularly strong slerotiza- tions (Fig. 25). The lateral basal mandibular margin is again more con• spicuous as a sclerotised mandibular ridge (smr. Fig. 25). It forms together with a large processus paratento• rialis the medial part of this anterior articulation com• plex. The processus paratentorialis clings to the medial side of the mandibular ridge (Fig. 33d). In this way the lateral clypeal border together with the processus para• tentorialis form a yoke-like cuticular framework of the Fig. 19. A. perscitus, left mandible with entire musculature cranium that encloses the lateral mandibular rim. The from posteroventral. processus paratentorialis does not extend across the lat- 160 A. H. STANICZEK erobasal rim of the mandible, but clearly is positioned ture, as the connecting membrane between basal man• medially to it. dible and cranium is located medially of the processus Apart from the lateral mandibular ledge, almost the en• paratentorialis. tire anterior articulation complex is not visible from outside, as it is overlapped by the protruding antennal Tentorium base and the antenna itself. Only when the antennal base is removed, a view of the entire articulation com• The anterior tentorial pit is overlapped by the antennal plex and the anterior tentorial pit is revealed. However, ridge and thus not visible externally. It is located below the processus paratentorialis is still an external struc- the antenna at the ventral border of the anterolateral

Fig. 22. T. gertschi, head in posterior view, postocciput and all mouthparts except of mandibles removed. M7 and M1c Fig. 20. T. gertschi, head in lateral view, antenna removed. are also removed to show the transverse ligament and the re• Circle shows a magnified schematic histological cut through maining portions of M1. postoccipital ridge.

Fig. 23. T. gertschi, head in posterior view. Postocciput. left posterior tentorial arm, M18 of right side, and all mouthparts except of left mandible removed to show the anterior tentorial Fig. 21. T. gertschi, head in lateral view, antenna and parts of arm and the cranial part of the anterior articulation complex. the lateral cranium removed to show the mandibular muscles. The left mandible is partly opened to show M5. The Mandible of Silverfish and Mayflies 161 cranium. The anterior tentorial arms medially split off nective muscle (M23, Figs. 23, 30). The posterior ten• a dorsal tentorial arm that is attached to the cranium by torial pit is not visible externally. Directly behind of tonofibrillae. Additionally, a tiny apodeme extends the mandibles, a conspicuous transverse fold crosses dorsad from the basis of the dorsal tentorial arm. The the head, that internally is represented by a massive cu• suspensor muscle of the tentorium (M18, Figs. 23, 30, ticular phragma (Figs. 20, 22-23). This transverse fold 33f) is attached to this apodeme. The anterior tentorial is regarded as homologue of the postoccipital suture arms converge posteriorly and turn into the tentorial herein (see discussion). When the specimen is dis• plate, which overlaps the posterior tentorial arms from sected from behind and the posterior head region be• anterior. The posterior tentorial arms however are well hind the phragma is removed, the posterior tentorial pit separated from the tentorial plate and only connected becomes apparent as inconspicuous elongated slit medially by a short muscular layer, the tentorial con• (Figs. 22-23).

Fig. 24. T. gertschi, right mandible with at• tached musculature in anterior view, SEM.

Fig. 25. T. gertschi, right mandible, anterior articulation complex, SEM. 162 A. H. STANICZEK

Mandibular muscles (Figs. 21-23, 28, 33) body is extended and forms a blunt-edged triangle that turns into the mandibular apodeme. Its tendon along a) Dorsal muscles with the entire inner basal border of the mandible forms an attachment for those portions of the M. cranio• (M1) M. cranio-mandibularis internus: mandibularis internus (la, 1c) that originate from the The cranial adductor muscle is the most voluminous ipsilateral vertex. A separate, smaller portion (lb) origi• mandibular muscle. The medial side of the mandibular nates from the head region posterior of the chitinuous

Fig. 26. T. gertschi. right mandible, posterior articulation, SEM.

Fig. 27. T. gertschi, mola and incisivi of right mandible from posterior, SEM. The Mandible of Silverfish and Mayflies 163 phragma and extends to the medial side of the mandi• (M3) M. cranio-mandibularis externus posterior: bular apodeme. This muscle consists of two portions that both insert with a common tendon on the basal outer mandibular (M2) M. cranio-mandibularis externus anterior: rim just below the compound eye. The anterior portion This relatively small and short abductor muscle has its has its origin on the lateral cranium directly above the origin immediately anterodorsal of the compound eye compound eye. The posterior portion originates on the on the lateral cranial wall. It extends anteroventrally lateral part of the chitinuous phragma. and inserts on a long and tenuous apodeme immedi• ately posterior to the posteromedial part of the anterior articulation complex. b) Ventral muscles

(M4) M. tentorio-mandibularis externus dorsalis: This flat muscle originates at the lateral brink of the pretentorium and inserts at the basal lateral border of the mandible between the anterior and posterior articu• lation.

(M5) M. tentorio-mandibularis externus ventralis: Almost the entire mandibular cavity is filled by this massive muscle, that extends medially. It originates with several portions from a broad ligamentous connec• tion that links the respective muscles of both sides of the head. The transverse ligament closely approaches the tentorium posteroventrally, and the cross-section reveals even some attachments between the two structures.

(M6) M. tentorio-mandibularis internus dorsalis: This small and inconspicuous muscle extends from the Fig. 28. T. gertschi, right mandible with musculature from lower third of the dorsal tentorial arm to its insertion at anterior. M7 is not visible. the dorsal posterolateral angle of the mandible.

(M7) M. tentorio-mandibularis internus ventralis: M7 is a flat muscle that originates from the ventral side of the tentorium where the anterior tentorial arm takes its rise from the corporotentorium. It inserts just ven- trally of the insertion of (M1) at the upper inner side of the mandibular cavity. It extends thus behind all other ventral muscles.

(M8) M. mandibulo-hypopharyngealis: The presence of a mandibular-hypopharyngeal muscle could neither be confirmed in a dissected specimen nor in a histological cross-cut section series.

Fig. 29. The development of the anterior condyle and the 3.5. Lepisma saccharina subgenal ridge in Dicondylia: Schematised cross-sections through the right mandible and cranium at the level of the The mandible in L. saccharina is very reminiscent of proc. paratentorialis in a.) Zygentoma (Tricholepidion gert• schi) b.) Ephemeroptera (Siphlonurus croaticus) c.) Meta- the condition in T. gertschi, but it differs in the reduced (Acheta domesticus). d.) Cross-section between an• pars molaris, the presence of M8, the disappearance of terior and posterior mandibular articulation to show the sub- the transventral ligament and consequent shift of M5 to gena in Metaptergota (Acheta domesticus). For further expla• the anterior tentorial arm (see also FÜRST V. LIEVEN nation see also text. 2000). 164 A. H. STANICZEK

4. DISCUSSION served in living as well as in fixated specimens, espe• cially when the entire cibarium was filled with col• 4.1. Functional aspects of the mandibular lected detritus. movement The mandibular notch of the anterior articulation com• plex is not able to perform an extensive pro- and retrac- 4.1.1. Zygentoma tional gliding along the genal margin, because it is in• hibited posteriorly by the processus paratentorialis and The anatomy of the posterior articulation allows mainly anteriorly by the transverse margin of the clypeus (Fig. two possible modes of mandibular movements: 5). Thus it seems not appropriate to term this structure First there is a gliding movement along the longitudi• as "slider" or "sliding articulation" as proposed by sev• nal axis of the posterior condyle, that would result in eral authors before (ARENS 1989; KUKALOVÁ-PECK an oblique gliding of the entire mandible in anterome- 1991). SCHÖNMANN (1979, 1981) and STRENGER (1953, dial or posterolateral direction. This movement would 1970, 1973, 1975, 1977, 1979) interpreted this conspic• mainly be caused by the action of the ventral muscles. uous mandibular hollow groove and its cranial counter• Recent observation on living specimens of Lepisma part as a device that has to withstand and absorb the saccharina by FÜRST V. LIEVEN (2000) establishes that forces (i.e. the pressure of the mola) of the respective this indeed is the prevailing mode of mandibular adducting contralateral mandible, and I completely movement during feeding. In this case the anterior ar• agree with this interpretation. ticulation complex solely serves as a guiding structure From its lowered position the mandible can be pro• for the mandible. Especially the processus paratentori• tracted until the anterior articulation complex gets in alis does not serve as a condyle, but only as a medial contact. This movement is obviously caused by the ac• guiding brace, that prevents the mandible from tilting tion of M2. Only when the anterior articulation com• inwards. plex is tightly connected, the change to an abductional/ Second a limited rotation around the longitudinal axis adductional movement can take place. Then the mandi• of the posterior articulation would result in a slight ab- bular notch of the anterolateral part rotates around the ductional and adductional movement. In this case the sclerotised genal margin, and the mandibular rim of mandible would rotate around an axis between the ante• posterolateral part rotates around the processus paraten• rior and posterior condyle. M3 would act abductor and torialis. The axis of rotation runs between the postero• M1 would act as adductor of the mandible. medial part of the anterior articulation and the posterior As the anterior articulation complex is just a loose con• articulation. nection, it needs additional tightening. This may the ex• During the more or less transversly orientated ab- and plain the insertion of M2 close to the posteromedial part adduction the condyle of the posterior joint describes of the articulation complex. a complicated path. When the abduction takes place, Regarding the loose anterior articulation complex as the roller first glides along its longitudinal axis pos- well as the possibility to perform also pro- and retracto- terolaterad. At the same time it rotates clockwise (at rial movements, it may be justified to term this kind of least the left mandible, the right mandible accordingly mandibular arrangement as facultative dicondyly. counterclockwise) around its socket, until the longitu• dinal axis of the roller lies in the frontal plane. Addi• 4.1.2. Ephemeroptera tionally the roller gets lifted laterally and lowered me• dially, until its longitudinal axis has reached the hori• • Oniscigaster wakefieldi zontal plane. Adducting the mandible, the posterior The main difference compared to the mandibular artic• condyle describes the same movements in opposite ulation of silverfish concerns the modified anterior ar• sequence. ticulation complex. Especially its posteromedial part This trace I have observed by manipulating the man• plays a different role in the movement of the mandible. dible in fixed specimens, and my own observation on The processus paratentorialis rests medially of the feeding Baetis and Siphlonurus (unpublished data) co• mandibular rim, but it also extends on top of it, thus incide, where I have found this movement to happen re• serving as a true secondary condyle. It is still not a tight peatedly with a high frequency. In addition. BROWN ball-and-socket joint, but just a loose attachment that (1961) observed in Baetidae (Ephemeroptera) also this can easily be disconnected when the mandible is low• rhythmic slide of the posterior articulation along its ered. In this case also the anterolateral part of the ante• longitudinal axis. rior articulation complex gets disconnected, and its an• The observed movements of the posterior articulation terior border forms an acute angle with the inflected should also imply some rotation of the entire mandible gena. The posterior articulation accordingly rotates around the processus paratentorialis. Such a movement around its longitudinal axis. This position could be ob• is most likely caused by the action of M7. The Mandible of Silverfish and Mayflies 165

• Nesameletus sp. mandible, which is without any doubt a huge selective The entire anatomy and function of the mandibular ar• advantage for a carnivorous insect. ticulation is basically the same as in O. wakefieldi. The Most intriguing are the very different points of attach• main difference concerns the shape and position of the ment of the anterior cranial abductor muscle, M2. The posterolateral part of the anterior articulation complex. anterior shift of its origin as well as the posterior shift In Nesameletus both parts of the anterior articulation of its insertion brings along a different direction of M2. complex are widely separated, because its posterome• that increases its protractional component. dial part is located at the posterior end of the anterior Accordingly the insertion of the posterior cranial ab• tentorial arm. This results in a position that lies more ductor M3 is now located anterior of the insertion of dorsal and thus also affects the rotational axis of the M2, which increases the refractional component of the mandibular swing. Another interesting feature is the former. The reinforced ability to perform re- and pro• structure of its posteromedial part. The mandibular rim tractional movements of the entire mandible goes ahead is elevated and forms a broad and flat plateau on which with the shape of the posterior condyle, whose longitu• the likewise flattened processus paratentorialis rests. dinal axis lies in the parasagittal plane. When the man• Possibly the rotation of the entire mandible around the dible is entirely adducted, the processus paratentorialis posterolateral part of the anterior articulation complex is located at the posterior end of its mandibular socket. is increased in Nesameletus sp. In abducted position the processus paratentorialis is lo• cated at the anterior end of the mandibular socket. It may be not the mandible that is initially used for seiz• • Ameletopsis perscitus ing the victim, but the enormously enlarged maxilla, The anatomy of the mandible is strongly modified. The whose galeolacinia is modified to a huge claw. The loss of the mola. the reduction of the ventral mandi• mandible may be used to cling to the prey and to engulf bular muscles, and the missing mandibular notch of the it into the pharynx during adduction. When it becomes anterior articulation complex are linked with the car• abducted, the incisivi must be retracted without pulling nivorous life habits. the victim out of the mouth again. In this context the The partes molares of O.wakefieldi are placed ventrally desclerotised strips at the basis of the incisivi could be of the morphological mouth opening. Here they com• of importance. They allow a passive jackknifing of the press the detritus that has been accumulated by maxil• incisivi, so they can easily be retracted and pass the lae and mandibular tips of incisivi. A. perscitus how• prey laterally without difficulty. This passive jackknif• ever engulfs entire insect larvae. Voluminous molae ing has not been observed in living specimens yet, but would only be obstructive to the prey's quick passage the ability to do so can easily be proven by manipulat• through the cibarium. ing the incisivi. In addition it is necessary to reach a great width of ab• duction for those mouthparts that seize the prey, cling to it, and engulf it into the pharynx. This requirement 4.2. The anterior mandibular articulation can only be fulfilled by a reduction of the tentorio-man- of Zygentoma and Ephemeroptera: dibular muscles that limit the range of abduction. This a brief history of misunderstandings may explain the loss of M4-M6 in A. perscitus. In O. wakefieldi the mandibular notch stabilises the HENNIG (1953) was not the first one who realised the mandible during feeding and enables the transfer of the presence of a dicondylous mandible in Zygentoma, but chewing pressure to the cranial side wall and the tento• he was the first to point out its phylogenetic signifi• rium. In A. perscitus no chewing pressure of the molae cance by establishing the monophyletic taxon Di- is effective, so the loss of the mandibular notch be• condylia. It has been BÖRNER (1909), who early recog• comes understandable. The missing chewing pressure nized the similarities between the "thysanuroid" man• also led to the reduction of the anterior tentorial arms dible of Zygentoma and Ephemeroptera on the one that serve as place of origin for the M7 and the proces• hand and the "orthopteroid" mandible of the remaining sus paratentorialis. This reduction - like the loss of a pterygote insects on the other hand, which he called common corporotentorium - might also be correlated Metapterygota. However, a detailed comparison of the with additional space required for the extremely expan• mandibular articulations as well as their muscles has sible pharynx. never been undertaken or failed, because species were The apodemes of all three cranio-mandibular muscles not chosen which were still close to the respective are very much shortened or even lost. Thus the muscles groundplans. grow in length compared to the condition in O.wake• Additionally, the authors who described the mandible fieldi or Nesameletus sp. As a result the muscles are of silverfish came to different conclusions. While all able to perform a much faster ad- or abduction of the authors agreed on the presence of an anterior articula- 166 A. H. STANICZEK tion, in detail their observation of this character differed sumption that, in addition to the posterior mandibular considerably. SNODGRASS (1950, 1952, 1960) repeat• articulation, an anterior articulation complex with two edly located in Lepisma and Ctenolepisma the anterior attachment points between cranium and mandible can condyle "on the ventrally inflected angle of the gena, indeed be attributed to the groundplan of larval ...a short distance behind the clypeus", CHAUDONNERET Ephemeroptera. Moreover, a comparison of the two dif• (1950) instead observed in Thermobia an anterior ferent character states of its posteromedial part to the condyle on the laterodorsal angle of the clypeus. In fact condition in silverfish clearly points to an anterocondy• both of them have been right and wrong. CHAUDON• late condition in the groundplan of Ephemeroptera. Be• NERET observed only the anterolateral part of the ante• sides O. wakefieldi I have found the plesiomorphic ante• rior articulation complex, while SNODGRASS took the rocondylate character state of the mandible in processus paratentorialis of the posteromedial part for Ametropodidae, Siphlonuridae, Ameletidae, Colo- the anterior condyle. Later it was SMITH (1988) who buriscidae, and Siphlaenigmatidae, whereas in all other briefly mentioned a "semipermanent" anterior man• investigated families the posterocondylate condition dible articulation in Zygentoma, and KUKALOVÁ-PECK was found (Nesameletidae, Baetidae, Heptageniidae, (1991), who for the first time referred to the anterior ar• Oligoneuriidae, Leptophlebiidae, all investigated ticulation of silverfish as "two sockets which fit into a Ephemeroidea and Pannota). In any case the shift from clypeotentorial yoke". KUKALOVÁ-PECK (1985, 1991) anterocondylate to posterocondylate mandible must however did not recognize the presence of the postero• have happened several times in the evolution of medial part of the anterior articulation in mayfly larvae. mayflies, as there are several distinct monophyla (for SNODGRASS (1950) recognized only the primary mandi• instance Baetoidea, see STANICZEK 1996, or Setisura, bular joint in the Ephemeroptera and thus assumed a see MCCAFFERTY 1991), that still have a basal antero• convergent development of the secondary joint in both condylate taxon (Siphlaenigmatidae, Coloburiscidae), Zygentoma and Metapterygota. STRENGER (1953, 1970, whereas other taxa inside this monophylum have 1975, 1977, 1979) and BROWN (1961) always observed aquired the posterocondylate condition (Baetidae, Hep• an anterior, middle, and posterior mandibular attach• tageniidae + Oligoneuriidae). ment to the cranium in various mayfly species through• out different families, but did not try to homologise these attachment points with the anterior (secondary) 4.3. The homology of the mandibular joints and posterior (primary) joints of other Dicondylia. throughout the Dicondylia However, SCHÖNMANN investigated Siphlonurus aesti• valis (Siphlonuridae), a mayfly that is thought to resem• There is a remarkable correspondence between the ble the mayfly groundplan in many aspects (SCHÖN• mandibular articulation in silverfish and mayflies. Both MANN 1981), but he could not confirm the presence of exhibit an anterior articulation complex, whose poste• the middle attachment in this species. He thought of the rior part is formed by a processus paratentorialis that is middle attachment of the ephemeropteran mandible as located immediately ventral to the anterior tentorial pit. homologue of the secondary mandibular joint of other These processus represent homologous structures, and Dicondylia, and assumed its reduction in Siphlonurus. thus should be attributed to the groundplan of Di• This study clearly confirms the results of STRENGER and condylia. An independent development of such a pecu• BROWN. The anterior attachment corresponds to the an• liar arrangement in Zygentoma and Ephemeroptera ap• terolateral part of the anterior articulation complex, and pears unlikely. The anterolateral part of the anterior ar• the middle attachment to its posteromedial part. Most ticulation complex is however somewhat different in interesting however is the fact that all species investi• mayflies and silverfish: In Zygentoma, including T. gated by STRENGER and BROWN exhibit the postero- gertschi, it is the lateral clypeal border that forms a cu- condylate condition. Here the processus paratentorialis ticular stiffening which makes contact with the mandi• is located well separated from the anterolateral part of bular groove. In mayfly larvae the mandibular groove is the anterior articulation complex and thus very easy to deeply excavated and encloses the inflected gena. How• distinguish as separate articulation. A closer look at ever, it seems likely to interpret the altered anterior ar• Siphlonurus aestivalis, the species investigated by ticulation complex of mayfly larvae as a modified ante• SCHÖNMANN, indeed revealed that this species shows rior part of the dicondylian articulation complex that the anterocondylate condition just as O. wakefieldi fits the specialized demands for the intake and com• does. SCHÖNMANN overlooked the posteromedial part of pressing of aquatic detritus. The alteration of the man• the anterior articulation, simply because it was absent dibular groove to a mandibular notch in mayfly larvae from that location where it has been previously found in and the shift of its cranial counterpart to the gena thus other species. My subsequent investigation of various represent autapomorphies of Ephemeroptera. A com• mayfly nymphs throughout the order leads to the as• mon origin of the anterolateral part of the anterior artic• The Mandible of Silverfish and Mayflies 167 ulation and its subsequent change in mayflies appears is an occipital ridge present. Thus BITSCH (1963) named more likely than its independent acquirement in both it occipital suture. Internally this suture is partly ex• Zygentoma and Ephemeroptera. tended into two large laterodorsal cranial apodemes, The homology of the posterior (primary) articulation that have been accordingly termed occipital phragmata. throughout the insects is obvious. It is more difficult to In Machilis the cranial remotor muscle of the mandible determine the precursor structure of the secondary ar• is attached to the medial part of this ridge, just medial ticulation, that is formed as ball-and-socket joint in the of the two laterodorsal phragmata. The phragmata are remaining pterygote insect orders. Generally, the points of insertion for the dorsal longitudinal muscles condyle of the anterior articulation is located at the of the prothorax. Medially of the phragmata the dorsal basal lateral margin of the clypeus in Odonata (ASA- mandibular remotors of the mandible (crm1 BITSCH HINA 1954; HAKIM 1964) as well as in the orthopteroid 1963) are attached to the ridge. LAUTERBACH (1972b) (SNODGRASS 1935; STRENGER 1942; WALKER 1931, realised that in Pterygota the prothoracic and neck mus• 1933) and other lower pterygoid orders (CRAMPTON cles are usually attached to the postoccipital ridge, and 1932; MOULINS 1968). In Zygentoma it is also the thus homologized the archaeognathan suture in ques• clypeal border that contributes to the anterolateral part tion with the postoccipital ridge of pterygote insects. of the anterior articulation complex. It might be tempt• Moreover, he also realised the plesiomorphic presence ing to homologise the anterolateral part with the sec• of small pleural folds in T. gertschi, that have been con• ondary condyle of the remaining pterygotes (see FÜRST served from the groundplan of Ectognatha (LAUTER• VON LIEVEN 2000). However, this would not only de• BACH 1972b). mand a secondary shift of the cranial part of the antero• The present study indeed confirms these homologisa- lateral joint to the gena in mayflies, but also an inde• tions, and specifies the account on T. gertschi in pendent development of the functional anterior articu• LAUTERBACH's (1972b) scenario on the evolution of the lation in mayflies and Metapterygota. There are more postoccipital ridge in Ectognatha: facts that do not concur with that hypothesis: first of all In T. gertschi there is a cranial transverse fold present the mandibular groove of the anterior articulation in (Fig. 20, circle) that is internally extended into a promi• Metapterygota is generally located at the basal rim of nent large and flattened ridge. It is not only present at the mandible, but in both Zygentoma and Ephemero• the laterodorsal part of the head, but it stretches as a ptera it is well separated hereof. It is in fact the proces• continuous apodeme across the entire laterodorsal cra• sus paratentorialis that articulates with the basal mandi• nium (Fig. 23). This extended ridge serves also as an at• bular rim. Second, it is the processus paratentorialis tachment area for the prothoracic muscles. Moreover, a that works as the functional anterior articulation in minor part of the cranial adductor muscle (M1) is at• mayfly larvae. It seems likely that a similar structure al• tached to the posterior side of the phragma. In Ephe• ready served in the groundplan of Pterygota as func• meroptera and other Pterygota also a minor part of M 1 tional condyle, from which the conditions in Ephe• originates from the postoccipital ridge. The overall sim• meroptera and Metapterygota derive. Otherwise an in• ilarity of the apodeme in and T. gertschi dependent development of different structures as ante• on the one hand and the matching muscular attach• rior joint in Ephemeroptera and Metapterygota had to ments of the postoccipital ridge of the remaining Zy• be assumed. Last but not least the anterior condyle in gentoma and Pterygota on the other hand clearly indi• Metapterygota is always associated with the position of cate their homology. The structure in question is indeed the tentorial arm (SNODGRASS 1935), which would homologous to the postoccipital ridge of other Di• again point to the posteromedial part of the anterior ar• condylia. ticulation complex as homologue to the anterior mandi• In the past there has been some confusion on the head bular articulation of Metapterygota. Accordingly, the sutures of T. gertschi: WYGODZINSKY (1961, fig. 6) anterolateral part of the anterior articulation must have marked the transverse fold erroneously as membrane• been lost in Odonata and . ous area, and SHAROV (1966, fig. 47B) followed him in this regard. A part of the postocciput itself is inhomoge- nously pigmented and heavily folded crosswise, so 4.4. The homology of the postoccipital ridge WYGODZINSKY (1961) took one of these folds posterior throughout Dicondylia to the postoccipital ridge for the latter. LAUTERBACH (1972b, fig. 7), obviously relying on these data, erro• In Archaeognatha there is a prominent suture that ex• neously homologised one of the posterior folds with the tends across the head from one posterior mandible postoccipital ridge. BOUDREAUX (1979b, fig. 52) took condyle to the other. The lateral ends of the suture are the postoccipital ridge for the occipital ridge, and the located slightly anterior to the primary mandibular fold on the postocciput itself for the postoccipital ridge. condyle. This is also the case in Pterygota, when there Nevertheless it has been LAUTERBACH (1972b), who Tab. 1. Synonyms of the mandibular muscles in Ectognatha.

This publication BITSCH(1963) CHAUDONNERET (1950) BÖRNER(1909) BROWN (1961) STRENGER SCHÖNMANN (1981) Archaeognatha Zygentoma: Zygentoma Ephemeroptera (1953-1977) Ephemeroptera Thermobia domestica Ephemeroptera Ephemeroptera

M. cranio-mandibularis muscle cranio- adducteur cranien médial Adductor mandibulae dorsal adductor dorsaler (cranialer) dorsaler internus mandibulaire 1 abducteur cranien médial posticus muscles Mandibeladduktor Mandibeladduktor Ml (crml) (adcm, abcmd) (a.m.p.) (6) (MA) (MAa-d)

M cranio-mandibularis muscle cranio- adducteur Remotor mandibulae cranial abductor _ cranialer externus anterior mandibulaire 2 cranien latéral anticus muscle Mandibelabduktor M2 (crm2) (adcl) (r.m.a.) (3a) (25a)

M. cranio-mandibularis muscle cranio- abducteur cranien Remotor mandibulae cranial abductor cranialer cranialer externus posterior mandibulaire 3 antérieur posticus muscle Mandibelabduktor Mandibelabduktor M3 (crm3) (abca) (r.m.p.) (3b) (25) (25b)

M. tentorio-mandibularis muscle tentorio- abducteur _ tentorial abductor tentorialer tentorialer externus dorsalis mandibulaire 2 tentorial latéral muscles Mandibelabduktor Mandibelabduktor M4 (ttm2) (abtl) (4) (in partim) (25') (25')

M. tentorio-mandibularis adducteur adducteur Adductor mandibulae tentorial adductor Adductor tentorialer externus ventralis mandibulaire 1 tentorial principal tentoricus muscles tentoricus Mandibeladduktor M5 (adml) (adt) (a.m.t.) (in partim) (5) (in partim) (26b, 26) (26a)

M. tentorio-mandibularis muscle tentorio- abducteur tentorial abductor Adductor tentorialer internus dorsalis mandibulaire 3 tentorial postérieur muscles tentoricus Mandibeladduktor M6 (ttm 3) (abtp) (4) (in partim) (26') (26c)

M. tentorio-mandibularis muscle tentorio- adducteur Adductor mandibulae tentorial adductor Adductor tentoricus internus ventralis mandibulaire 1 tentorial accessoire tentoricus muscles tentorialer Mandibeladduktor M7 (ttm 1) (adt') (a.m.t.) (in partim) (5) (in partim) (26, 26a) (26b)

M. mandibulo- adducteur adducteur hypopharyngeal- hypopharyngealis mandibulaire 2 hypopharyngien Zungenmuskel mandibular muscle M8 (adm 2) ? (adh) (zgm.) (7) (29) (29) The Mandible of Silverfish and Mayflies 169

pointed out for the first time the distinct reduction of as an apomorphic character state to the groundplan of the postocciput along with the cranial pleural folds in Metapterygota. the remaining Dicondylia. Taking this into account, the Furthermore, the anterior tentorial arm in Ephemero• reduced size of the postocciput as well as the loss of ptera has also a significantly posterior position com• pleural folds may indeed be putative synapomorphics pared to the anterior tentorial arms of Zygentoma, of (Lepismatidae + Nicoletiidae) and Pterygota. Odonata, and also of the orthopteroid insect orders (HUDSON 1945, 1946). Anterior to the anterior tentorial pit the mandibular notch of the anterior articulation 4.5. The anterior tentorial arms throughout complex is located, so the development of the massive Dicondylia notch is probably responsible for the back shift of the anterior tentorial arm. This posterior position of the an• Compared to other Dicondylia, SNODGRASS (1951, terior tentorial arm is considered as an autapomorphy 1952) noticed a different position of the anterior tento• of Ephemeroptera (STANICZEK 1996). as in Odonata rial pits of both Zygentoma and Ephemeroptera in re• there seems to be the same condition as in orthopteroid gard to the lateral cranial wall. As shown in this study, orders (SNODGRASS 1928, 1955; SHORT 1955). the anterior tentorial pit in both groups is not visible ex• ternally, because it is located at the basal sclerotized margin of the inflected lateral cranium. But, contrary to 4.6. The homology of the mandibular SNODGRASS (1951), this does not imply a location musculature throughout Dicondylia "...mesad of the mandibles". In fact, the anterior tento• rial pit along with the processus paratentorialis is lo• It is surprisingly easy to compare the mandibular mus• cated immediately dorsal to the laterobasal rim of the culature of Archaeognatha (as documented by Bitsch mandible, as the laterobasal part of the mandible in 1963), Zygentoma, and Ephemeroptera and to trace its both Ephemeroptera and Zygentoma is likewise curved fate up to the Metapterygota. Fig. 30 summarizes the inwards. homologies of the mandibular muscles as seen in this In the basal orthopteroid lineages the anterior tentorial contribution, and Tab.1 lists the terminology of previ• pits are generally present as elongated slits in both the ous authors. SNODGRASS (1935, 1950) differentiated epistomal and subgenal sutures, the anterior condyle just one pair of dorsal and two pairs of ventral muscles being located at the very corner where the lateral in ectognath insects. According to him, the anterior pair elypeal margin bends over to the pleurostoma. In of the ventral muscles would be the homologue to the Odonata the anterior tentorial pits are also located in the mandibulo-hypopharyngeal muscles of higher ptery• subgenal ridge (SNODGRASS 1928). gote insects. However, the actual situation is more com• These differences in the location of the anterior tento• plex: rial pit along with its associated condyle led TUXEN In Archaeognatha there is one big remotor muscle (1970, fide KRISTENSEN 1975) to believe in an indepen• (crm1, BITSCH 1963), that extends with a long apodem dent development of the anterior mandibular articula• from the occiput to the posterior mandibular rim. This tion. In fact it is more plausible and parsimonious to as• muscle can be homologized with M1, the cranial ad• sume a simple lateral shift of the anterior tentorial arm ductor of the mandible. There are also two cranial pro- in Metapterygota. This went along with a strengthening motors in Archaeognatha (crm1, crm2, BITSCH 1963), of the lateral wall that resulted in the development of both inserting at the anterior basal rim of the mandible. the subgenal ridge. LAUTERBACH (1972b) assumes the They clearly correspond to M2 and M3, the anterior development of a subgenal ridge already in the ground- and posterior cranial abductors. Both muscles are pre• plan of Dicondylia. However, in all investigated Zygen• sent in all investigated Zygentoma and some of the in• toma as well as in all investigated Ephemeroptera I vestigated Ephemeroptera, and thus can be attributed to have not found a prominent internal ridge like the one the groundplan of Ectognatha. M2 is certainly present that is developed in the orthopteroid insects for instance in the groundplan of Ephemeroptera, but obviously has (contra CHAUDONNERET 1950 and STRENGER 1952). been lost in many mayfly families. The loss of M2 in Apparently it is the lateral shift of the anterior tentorial Ephemeridae, Palingeniidae, Euthyplociidae, and Hep- arm, that goes along with the development of a true tageniidae (STRENGER 1953-1977) is probably corre• subgenal ridge (Fig. 29). Additionally, a broadening of lated with a tightening of the anterolateral part of the the lateral part of the anterior tentorial arms took part to anterior articulation complex and a further immobiliz• counterpart the enhanced forces during feeding. How• ing of the mandible (ARENS 1989). In all other ptery• ever, this character complex is not properly investigated gote lineages, there is only one cranial abductor pre• throughout Odonata. but it might well be possible, that sent, that usually originates from the postoccipital the development of a subgenal ridge can be attributed ridge. Hence it is homologized with M3. M2 obviously

i 170 A. H. STANICZEK

6: Metapterygota: alp of aac as well as M2 lost, laterodorsal shift of ata and pp, so axis of mandibular movement is anteriorly lifted, pmp as well as posterior ar• ticulation form tight ball-and-socket joints, M4 and M6 lost, M7 reduced 5: Ephemeroptera: alp of aac modified into mandibular notch to withdraw chewing pressure, ata shifted posteriorly 4: Pterygota: fusion of ata with pta, loss of M18 and M23, processus paratentorialis serves as functional anterior condyle, reduction of M5 3: loss of zygomatic ligament, consequent shift of M5 to anterior tentorial arm; M5 splits off a mandibulo-hypopharyngeal muscle M8 2: Dicondylia: anterior tentorial arms united, M23 present, enlargement of M1, aquisition of an anterior articulation complex, posterior condyle as gliding device 1: Ectognatha: posterior tentorium present 30

Fig. 30. Phylogenetic argumentation scheme of Ectognatha, only characters regarding mandible and tentorium are listed. For further explanation see also text.

is lost in both Odonata and Neoptera, which is probably associated muscle (M5) is still unaltered in T. gertschi, correlated with the fixed anterior articulation, too. and the respective muscle in Archaeognatha (adm) is BÖRNER (1909) assumed the cranial abductor muscle of without any doubt homologous to the latter. In (Lepis- Locusta (Caelifera) to be homologue of both M2 and matidae + Nicoletiidae) on the one hand and Pterygota M3 of the mayfly larva, because it has two portions, on the other hand the transverse ligament is lost and the one of which originates at the gena, the other at the pos• origin of M5 has shifted to the anterior tentorial arm. toccipital ridge. However, in the groundplan of Di• Additionally, a small mandibulo-hypopharyngeal mus• condylia M2 also has two portions, and this condition is cle (M8) is present in both taxa. Previous authors (e.g. also still preserved in Zygentoma. Therefore it seems SNODGRASS 1935) homologized M8 of Pterygota with more likely that the bipartite cranial abductor of Caelif• the ventral adductor muscle adm of Archaeognatha, but era is rather the homologue of M2 only. the comparison of T. gertschi with other Zygentoma In Archaeognatha there is a big ventral adductor muscle rather points to an origin of M8 as split off from M5. In extending from the mandible to the transverse medial my view it is most likely that the mandibulo-hypopha• ligament (adm1-adm5) (BITSCH 1963). Though de• ryngeal muscle M8 is primarily missing in the ground- scribed as several separate muscles by BITSCH, I have plan of both Ectognatha and Dicondylia, and the char• not been able to verify these separated portions in either acter distribution obviously points to the aquisition of a histological cross-sections or dissected specimens of mandibulo-hypopharyngeal muscle as an apomorphic Petrobius brevistylis. The transverse ligament and the character state of (Lepismatidae + Nicoletiidae) + The Mandible of Silverfish and Mayflies 171

Pterygota. Additionally, the tendency to reduce M5 and M8 is often regarded as a possible autapomorphy of Metapterygota (Kristensen 1998), although in dragon- (in larvae and adults) both muscles are still rather well developed (Snodgrass 1955; Short 1955). In Archaeognatha there are three tentorial muscles, that extend from the anterior (ttm2, ttm3) and posterior (ttml) mandibular rim to the anterior tentorial arm (Bitsch 1963). The latter muscle is easy to homologise with M7 in T. gertschi, but due to the enormous en• largement of the insertion of M5 at the medial side of the mandibular rim, the insertion of M7 is shifted ven- trally to the mandibular body in the groundplan of Di• condylia. The derivation of M6 is not obvious. It is present at the medial mandibular rim of all investigated Zygentoma and Ephemeroptera. It may represent a split off from ttml, but more parsimonious is the assumption of a ho• mology with ttm3 of Archaeognatha. However, this im• plies the shift of its insertion to the medial side of the mandibular rim in the groundplan of Dicondylia. This leaves ttm2 as possible homologous muscle to M4, that is present as distinct layer with identical course in all investigated Zygentoma and Ephemeroptera. Apart from all these minor uncertainties it becomes ob• Fig. 31. P. brevistylis, histological cross-section through the vious, that in the groundplan of Dicondylia there is a head at the level of the partes molares. Four black arrows in• distinct superficial layer of muscles present, that ex• dicate the inner sclerotised ridges of the mandible. tends from the anterior tentorial arm to the basal mandi• bular rim. This distinctive muscle layer is entirely lost in Odonata and Neoptera, which may count for another putative autapomorphy of Metapterygota. the respective anterior tentorial arm could indeed be taken as arguments for a sistergroup relationship of (Lepismatidae+Nicoletiidae) with Pterygota rather than 4.7. Do the Zygentoma including Tricholepi- Lepidothrichidae. The reduction of the postocciput as dion gertschi represent a monophylum? well as of the pleural folds in these taxa might also be added to the above mentioned characters as putative WYGODZINSKY (1961) described the western Nearctic synapomorphies. relic species T. gertschi, and assigned it to the taxon Zy• On the other hand KRISTENSEN (1998) discusses, be• gentoma. However, T. gertschi lacks many of the char• sides the reduction of superlinguae, another two charac• acters that are generally referred to as autapomorphies ters as possible autapomorphies of Zygentoma, namely of Zygentoma or even Dicondylia (see KRISTENSEN sperm conjugation and shape of the apical segment of 1991). This would leave T. gertschi the sistergroup of the labial palp. However, a broadened distal labial palp the remaining Dicondylia. The results of the present in• segment occurs in a variety of mayfly larvae, and is vestigation indeed confirm the studies of BOUDREAUX also commonly seen throughout different insect orders, (1979a) and KRISTENSEN (1998) regarding the presence so I do not consider this as strong evidence for the of a ligamentous connection between the ventral man• monophyly of Zygentoma. Previously not discussed dibular adductor muscles M5, which consequently has phylogenetically is another peculiar feature of the Zy- to be assumed in the groundplan of Dicondylia. In con• gentoman labial palp: it bears four segments (CHAU• trast to all other Zygentoma as well as Pterygota, the DONNERET 1950; PACLT 1963, 1967). Apart from the mandibulo-hypopharyngeal muscle is missing in T. distal one all segments are also equipped with intrinsic gertschi. This muscle is also absent in the myriapod and muscles (CHAUDONNERET 1950). In the groundplan of entognath lineages, and to my knowledge absent in Ar• Pterygota the labial palp is made up of three segments chaeognatha as well (contra BITSCH 1963). So the split• only. In both Archaeognatha and the entognath orders ting off from the mandibulo-hypopharyngeal muscle the labial palps never exceed three segments as well. So M8 of M5 as well as the latter's separation and shift to one might a priori think of the four-segmented labial 172 A. H. STANICZEK

Fig 32a-g. O. wakefieldi, histological cross-sections through the larval head from anterior to posterior at dif• ferent levels of the mandible. The mandible has been cut tangentially. The Mandible of Silverfish and Mayflies 173

Fig 33a-f. T. gertschi, histological cross-sections through the head from anterior to posterior at different levels of the man• dible. The mandible has been cut tangentially.

palp of Zygentoma as an autapomorphic character state. verfish (STURM 1987a)- deposit the spermatophore on a On the other hand in Chilopoda a four-segmented pal• self-spun web rather than on a peduncle like the Entog- pus on the second maxilla is exhibited. It seems that the natha and Archaeognatha (STURM 1987b). At first this condition in Zygentoma could as well be interpreted as behaviour could be evaluated as an autapomorphic the preserved tracheate groundplan in this respect. character of the Zygentoma. However it might as well STURM (1997) investigated the mating behaviour of T. be attributed to the Dicondylian groundplan. Similarily gertschi, whose males -as well as the males of other sil• the presence of the tentorial connective muscle M23 174 A. H. STANICZEK

could be interpreted either as autapomorphy of Zygen• bly acts as protractor of the hypopharynx. The postoc• toma, as well as a character aquired in the stemline of ciput is significantly reduced. The mandibular trans• Dicondylia, that has been reduced in Pterygota due to verse ligament is lost, and the ventral adductor muscle the fusion of anterior and posterior tentorial arms. M5 becomes attached to the anterior tentorial arm. This At present the most parsimonious assumption would be enables the mandibles to move separately from each the monophyly of Zygentoma. but it would only need other. Otherwise the pattern of movement as well as the few opposite characters or a different character weight• anatomy of the mandibular articulation in the ground- ing to assume a sistergroup relationship between T. plan of (Lepismatidae + Nicoletiidae) remains unal• gertschi and all other Dicondylia. Obviously the pre• tered. sent data are contradictory and still too scarce to decide 4) In the groundplan of Pterygota the processus para• with a considerable amount of certainty if the Zygen• tentorialis extends onto the mandibular rim and starts to toma represent a monophylum or not. serve as true mandibular condyle. The anterolateral part of the anterior articulation complex is still preserved. There is still a pro- and retractional gliding of the entire 4.8. The transformation from monocondylous mandible possible, but the ab- and adductional compo• to dicondylous mandible, and the early nent of the mandibular movement increases. This is re- branchings of Dicondylia flected in the beginning reduction of M5 and further en• largement of M1. The increasing transverse forces lead Fig. 30 summarizes the gradual changes in tentorium, to another reinforcement of the inner skeletal frame, as mandibular musculature and modes of articulation in both anterior and posterior tentorial arms fuse to form ectognath insects, and the probable phylogenetic the characteristic pterygote tentorium. Consequently branchings as implied by the distribution of these char• the suspensor muscles of the anterior tentorial arms acters: (M18) and possibly also M23 are lost. 1) The monocondylous mandible of Archaeognatha is 5) In the groundplan of Ephemeroptera the pro-and re• able to swing in multiple planes. It bears several groups tractional movement is diminished by the alteration of of muscles: two cranial promotor muscles (M2, M3). the anterolateral part of the anterior articulation com• one cranial remotor muscle (Ml). two tentorial adduc• plex to form a deep mandibular notch, into which the tor muscles (M4, M6), and one transversely connected genal margin fits. The anterior tentorial arms have adductor muscle (M5) which move the mandible. With shifted posteriorly to make room for the enlarged pos• respect to the character distribution in entognathous in• teromedial part of the anterior articulation complex, sects (TUXEN 1959), this condition is proposed for the that is still close together in an anterocondylate condi• groundplan of Ectognatha as well. tion. The anterior cranial abductor muscle is retained. 2) In the groundplan of Dicondylia, as preserved in T. 6) In the groundplan of Metapterygota both articula• gertschi, the mandible still retains the transverse liga• tions become fixed into tight ball-and-socket-types of ment that links the ventral tentorial adductor mucles articulations. The axis of mandibular swing is about M5. All other muscles are retained, too. The lack of the horizontal. The anterior cranial abductor muscle M2 mandibulo-hypopharyngeal muscle is also a plesiomor- and the tentorial abductor muscles M4, M6 and M7 are phic trait shared with Archaeognatha. In contrast to the lost. The anterior tentorial arms along with the anterior latter, the mandible in the groundplan of Dicondylia has condyle shift laterally. A subgenal ridge is developed to aquired an additional articulation complex anterior of sustain the enhanced chewing pressure. The change to the primary joint that restricts the movements of the en• an obligatory dicondyle movement has taken place. tire mandible and serves as a guiding mechanism for the movements of the mandible. The anterior promotor M2 All of the investigated mandibular characters under• now acts as a stabilizer of the anterior articulation com• score the plesiomorphic character state of the mandible plex. The cranial adductor muscle Ml gets significantly of Zygentoma and Ephemeroptera in many respects, enlarged due to the different mode of movement. The and on the other hand point to a sistergroup relationship entire mandible is able to perform anteromediad respec• between Odonata and Neoptera with possible several tively posterolaterad orientated pro- and retractional synapomorphic character states: movements as well as an ab- and adductional move• - acquisition of a tight ball-and-socket-type of articu• ment. This is possible by the means of an elongated pos• lation, connected with the loss of the ability to protract terior condyle and the anterior articulation complex. and retract the entire mandible, broadening of the ante• Due to the increased transverse forces that occur during rior tentorial arm chewing, the anterior tentorial arms become united. - dorsal shift of the anterior articulation which is cor• 3) Lepismatidae+Nicoletiidae (and Pterygota) aquired related with the anterior lifting of the mandibular axis a hypopharyngeo-mandibular muscle M8. that proba• of rotation

• The Mandible of Silverfish and Mayflies 175

- acquisition of a subgenal suture and subsequent lat• there are still different views about the composition of eral shift of the anterior tentorial arm along with the the mandible in the groundplan of the Mandibulata and processus paratentorialis the groundplans of its subordinate groups: - loss of the anterolateral part of the anterior articula• Some authors assume a tripartite mandible in the tion complex groundplan of the Tracheata, which is interpreted either - loss of the anterior mandibular abductor muscle, M2 as an entire telopodite (KRAUS & KRAUS 1994, 1996) or - loss of the dorsal tentorial abductor muscles, M4 and as a coxopodite composed of three podites (KRAUS M6 1998; KUKALOVÁ-PECK 1998). These authors refer also - loss of tentorial adductor M7 to the mandibles of Archaeognatha. Zygentoma, or even larval Ephemeroptera (KUKALOVÁ-PECK 1998) as seg• This entire character complex thus can be added to the mented or at least showing traces of segmentation. previously suggested autapomorphies of Metapterygota However, the data obtained from this study do not sup• as repeatedly summarized by KRISTENSEN (1975, 1981, port this point of view. The existence of alleged "sutures 1991, 1998), such as the supression of imaginal moults, and sulci" that, according to KUKALOVÁ-PECK (1998), the presence of occlusor muscles in abdominal stig• occur in the mandible of mayflies, could not be con• mata, the reduction of the terminal filament, and the tra- firmed. It seems that KUKALOVÁ-PECK overinterpreted cheation pattern of wings and legs. Molecular data also the anterolateral and ventrolateral edges and angles of support a sistergroup relationship between Ephemero• the mayfly mandible or even the externally visible bor• ptera and Metapterygota (WHEELER 1989). ders of respective muscle attachments as remnants of On the other hand KUKALOVÁ-PECK (1985, 1991) as• segment borders. In my opinion the conditions that are sumed a sistergroup relationship between Ephemero• actually present do not justify these conclusions. ptera and Odonata. mainly by means of fossil wing Similarily, throughout all investigated species of Zy• characters. Wing characters may be difficult to polar• gentoma. including T. gertschi, no traces or remnants of ize, because the outgroups have no wings at all. The fu• a segmentation of the mandible could be demonstrated. sion of galea and lacinia is also often cited as a putative These findings remain in conflict with the postulated synapomorphic character of , but STAN- traces of segmentation in the mandible of Zygentoma ICZEK (2000) discusses the possibility that the fused by KRAUS & KRAUS (1994) and KRAUS (1998), that parts of the maxilla in Odonata may not represent galea however have not been documented. CHAUDONNERET and lacinia at all, but lacinia and palpus maxillaris in• (1950) in his detailed study on the firebrat Thermobia stead. However, the assumption of a sistergroup rela• domestica did not observe an external segmentation in tionship between Ephemeroptera and Odonata implies its mandible either, but nevertheless suggested a com• an independent development of a ball-and-socket type posite mandible made up of three segments, his only of mandible along with the associated changes in mus• evidence being the splitting of the mandibular nerve culature and tentorium in both Odonata and Neoptera. I into three main motoric branches. Whether this actually do not consider this to be likely. reflects a postulated former partition remains doubtful, as usually all main nerves tend to split when they have to supply different muscles. 4.9. Gnathobasic versus telognathic insect KRAUS & KRAUS (1994, 1996) observed in the machilid mandible Trigoniophthalmus alternatus Silvestri, 1904 "vestiges of original articulations between composing segments". The transformation of the first postantennal appendage Also KUKALOVÁ-PECK (1998) interpreted externally vis• into mandibles in crustaceans, myriapods and insects ible sutures in the mandibles of Archaeognatha as rem• has always been a major argument for a common origin nants of podite borders. In fact, a look onto a histologi• of these groups, which consequently led to their becom• cal cross-section of the mandible of Petrobius bre- ing the monophyletic taxon Mandibulata (SNODGRASS vistylis reveals that these sutures refer to heavily sclero• 1938; LAUTERBACH 1980). However, MANTON (1964) tised sulci. The same condition as shown in Fig. 31 has postulated an independent development of the man• been verified earlier also in Petrobius maritimus by dible in Crustacea and Tracheata. In contrast to the ob• MANTON (1964) and in several other machilids by viously gnathobasic mandible of crustaceans, MANTON BITSCH (1963). These sulci mark the proximal and dis• claimed a so-called "whole-limb jaw" in the ground- tal borders of the ventral muscle attachments. Obvi• plan of the Tracheata that is assumed to be made up of ously these inner ridges serve as devices that were sec• the entire telopodite. ondarily developed in order to strengthen the entire While there are also other arguments besides the man• mandible against forces that are effective during feed• dible for upholding the Mandibulata concept (for a ing, rather than indicating vestigial borders of leg comprehensive review of this topic see WÄGELE 1993), podomeres. Finally an alleged tripartite mandible in 176 A. H. STANICZEK

Diplura (MARCUS 1951) has never been confirmed by very much indepted to Dipl-Biol. A. Fürst von Lieven (Freie other workers on that group. Universitat Berlin, Germany), who not only commented on All in all, the assumption of a telognathic mandible in this work, but also kindly provided insights into his the groundplan of insects currently is not substantiated manuscript and frankly made his video tapes available to me. by morphological data. In addition to that, molecular He also waited patiently for the final version of my draft to allow a simultaneous publication. I am obliged to Dr. G. H. P. data, especially the mode of expression of the home- Bechly, Dipl-Biol. P. Bernstein, Dipl-Biol. A. Grau (all Uni• obox gene distalless (dll), point to a gnathobasic man• versität Tübingen, Germany), Mr. T. R. Hitchings (Canter• dible in the groundplan of insects (POPADIC et al. 1996), bury Museum, Christchurch, New Zealand). Dr. N. Ju. Kluge and due to recent studies by SCHOLTZ et al. (1998) even (University of St Petersburg, Russia), Prof. Dr. N. P. Kris- in the groundplans of all mandibulate lineages. tensen (Zoological Museum, University of Copenhagen, Den• KUKALOVÁ-PECK's (1998) hypothesis of a tripartite mark), and Prof. Dr. H. Sturm (University of Hildesheim, mandible in the groundplan of Crustacea is not substan• Germany) for loan or donation of material that enabled this tiated by the evidence obtained from both recent and study. I am also very grateful to Mr. T. R. Hitchings (Christ- fossil material: based on the comparison of recent church, NZ) and to one of the reviewers for their help with the species, LAUTERBACH (1972a) convincingly derived the English. Last but not least I thank Mr. H. Schoppmann who origin of the crustacean mandible not from the entire helped with the scanning electron micrographs. The Dept. of Conservation (Invercargill, New Zealand) permitted the col• protopodite, but only from its basal part, namely the lection of A. perscitus and Nesameletus sp. at Stewart Island. coxopodite along with its endite, as still can be ob• NZ (permit #93/5/8). This research was supported by the Lan- served in the larval development of many basal crus• desgraduiertenförderung Baden-Württemberg. taceans today. The gnathal lobe is always formed by the endite of the enlarged coxopodite (SNODGRASS 1952). In addition to that, WALOSSEK & MÜLLER (1990, 1998) References discuss the appendages of several fossils from the Orsten fauna which were assigned to the stem ARENS, W. (1989): Comparative functional morphology of lineage of Crustacea. In these fossils, WALOSSEK & the mouthparts of stream feeding on epilithic MÜLLER draw the attention to a separate endite proxi• algae. Arch. Hydrobiol. Suppl. 83: 253-354. mal to the undivided limb base, the latter carrying both ASAHINA, S. (1954): A morphological study of a relic dragon• exopodite and endopodite. Thus, according to its posi• Epiophlebia superstes Selys (Odonata. Anisozygo- tion the authors convincingly homologize this proximal ptera). 153 pp., The Japan Society for the Promotion of podite in the first postantennal segment with the cox• Science, Tokyo. opodite, hence the mandible. But WALOSSEK & MÜLLER BITSCH, J. (1963): Morphologie céphalique des Machilides evaluate the evolutionary aquisition of the coxopodite (Insecta ). Ann. Sci. Nat. 12: 501-706. as an autapomorphic trait of Crustacea only, and omit to BÖRNER, C. (1909): Neue Homologien zwischen Crustaceen und Hexapoden. Die Beißmandibel der Insekten und ihre discuss the consequences of their findings in regard to phylogenetische Bedeutung. Archi- and Metapterygota. the Mandibulata concept. If we accept the homology of Zool. Anz. 34: 100-125. the crustacean and the insect mandible, and this seems BOUDREAUX, H. B. (1979a): Significance of Intersegmental to be straightforward due to their identical body posi• Tendon System in Phylogeny and a Mono- tion as well as their structural similarities as modified phyletic Classification of Arthropoda. Pp. 551-586 in: appendages of the first postantennal segment, then GUPTA, A. P. (ed.) Arthropod phylogeny. Van Nostrand these Cambrian fossils indeed could be as well taken as Reinhold Company. representatives of the stem lineage of Mandibulata. In BOUDREAUX, H. B. (1979b): Arthropod phylogeny with spe• fact, this has already been suspected by LAUTERBACH cial reference to insects, viii + 320 pp.. John Wiley & Sons. (1988). New York. BROWN, D. S. (1961): The morphology and functioning of the Taking all known facts into account, then there is cur• mouthparts of Chloeon dipterum L. and Baetis rhodani rently little doubt that the mandible in the groundplan Pictet (Insecta, Ephemeroptera). Proc. Zool. Soc. Lond. of Mandibulata is a gnathobasic one, that is basically 136: 147-176. formed by a single podite, namely the enlarged cox• CAMPBELL, I. C. (1985): Dietary habits of Australian opodite along with its endite. In the groundplan of Tra- siphlonurid and oligoneuriid ephemeropteran nymphs. cheata the telopodite becomes lost. This view however Verh. Internat. Verein. Limnol. 22: 3250-3259. implies a secondary subdivision of the mandible in the CHAUDONNERET, J. (1950): La morphologie cephalique de myriapod lineages. (Packard) (Insecte Apterygote Thysanoure). Ann. Sci. Nat. 11: 145-302. CRAMPTON, G. C. (1932): A phylogenetic study of the head Acknowledgements. I am most grateful to Dr. G. Mickoleit capsule in certain orthopteroid, Psocoid, Hemipteroid and and Dr. E. Weber (Universität Tübingen, Germany) for dis• holometabolous insects. Bull. Brooklyn Entomol. Soc. 27: cussions and helpful comments on the manuscript. I am also 19-55. The Mandible of Silverfish and Mayflies 177

FÜRST VON LIEVEN, A. (2000): The transformation from LAUTERBACH, K. E. (1980): Schlüsselereignisse in der Evolu• monocondylous to dicondylous mandibles in the Insecta. tion des Grundplans der Mandibulata (Arthropoda). Abh. Zool. Anz. 239: 139-146. naturwiss. Ver. Hamburg (NF) 23: 105-161. HAKIM, Z. M. (1964): Comparative anatomy of the head cap• LAUTERBACH, K. E. (1988): Zur Position angeblicher Crus• sules of adult Odonata. Ann. Entomol. Soc. Am. 57: tacea aus dem Ober-Kambrium im Phylogenetischen Sys• 267-278. tem der Mandibulata (Arthropoda). Abh. naturwiss. Ver. HENNIG, W. (1953): Kritische Bemerkungen zum phylogene- Hamburg (NF) 30: 409-167. tischen Sytem der Insekten. Beitr. Ent. 3: 1-85. MANTON, S. M. (1964): Mandibular mechanisms and the evo• HUDSON, G. B. (1945): A study of the tentorium in some lution of . Philos. Transact. R. Soc. London, Ser. orthopteroid . J. Entomol. Soc. South. Afr. 7: B, Biol. Sci. 247: 1-183. 71-90. MARCUS, H. (1951): Observaciones morfologicas en Din- HUDSON, G. B. (1946): Studies in the comparative anatomy japyx marcusi. Folia Universitaria, Cochabamba 5: and systematic importance of the hexapod tentorium. II. 83-106. Dermaptera, and Isoptera. J. Entomol. Soc. McCAFFERTY, W. P. (1991): The Cladistics, Classification, South. Afr. 8: 99-110. and Evolution of the Heptagenioidea (Ephemeroptera). Pp. KLUGE, N. J., STUDEMANN, D., LANDOLT, P., & GONSER, T. 87-102 in: ALBA-TERCEDOR, J. & SANCHEZ-ORTEGA, A. (1995): A reclassification of Siphlonuroidea (Ephemero• (eds.): Overview and Strategies of Ephemeroptera and Ple- ptera). Mitt. Schweiz. Entomol. Ges. 68: 103-132. coptera, The Sandhill Crane Press, Gainesville, Fla. KRAUS, O. (1998): Phylogenetic relationships between higher MCLACHLAN, R. (1873): Oniscigaster wakefieldi, a New taxa of tracheate arthropods. Pp. 295-303 in: Fortey, R. Genus and Species of Ephemeridae from New Zealand. A. & Thomas, R. H. (eds.) Arthropod Relationships, Sys- Entomol. Mon. Mag. 10: 108-110. tematics Association spec. Vol. Ser. 55, Chapman & Hall, MCLEAN, J. A. (1970): Studies on the larva of Oniscigaster London. wakefieldi (Ephemeroptera: Siphlonuridae) in Waitakere KRAUS, O. & KRAUS, M. (1994): Phylogenetic system of the Stream, Auckland. N.Z. J. Mar. Freshw. Res. 4: 36-45. Tracheata (Mandibulata): On "Myriapoda": Insecta inter• MOULINS, M. (1968): Contribution a la conaissance ana- relationships, phylogenetic age and primary ecological tomique des Plécoptères: la région céphalique de la larve niches. Verh. naturwiss. Ver. Hamburg (NF) 34: 5-31. de Nemoura cinerea (Nemouridae). Ann. Soc. Ent. Fr. KRAUS, O. & KRAUS, M. (1996): On Myriapod / Insect Interre• (N.S.) 4: 91-143. lationships. Mém. Mus. natn. Hist. nat., Paris 169: 283-290. NICHOLS, S. W., TULLOCH, G. S. & TORRE-BUENO, J. R. de la KRISTENSEN, N. P. (1975): The phylogeny of hexapod "or• (1989): The Torre-Bueno glossary of entomology. 840 pp., ders". A critical review of recent accounts. Z. zool. Syst. The New York Entomological Society, New York. Evolut.-forsch. 13: 1—44. PACLT, J. (1963): Thysanura. Fam. Nicoletiidae. Genera In- KRISTENSEN, N. P. (1981): Phylogeny of Insect Orders. Ann. sectorum 216: 1-58. Rev. Entomol. 26: 135-157. PACLT, J. (1967):Thysanura. Fam. Lepidotrichidae, Main- KRISTENSEN, N. P. (1991): Phylogeny of Extant Hexapods. droniidae, Lepismatidae. Genera Insectorum 218: 1-86. Pp. 125-140 in: CSIRO (ed.) The Insects of Australia, 2n PENNIKET, J. G. (1962): Notes on New Zealand Ephemero• ed.,, Melbourne University Press, Melbourne. ptera. II. A preliminary account of Oniscigaster wakefieldi KRISTENSEN, N. P. (1998): The groundplan and basal diversifi• McLachlan, recently rediscovered (Siphlonuridae). Rec. cation of the hexapods. Pp. 281-293 in: Fortey, R. A. & Canterbury Mus. 7: 375-388. Thomas, R. H. (eds.) Arthropod Relationships, Systematics POPADIC, A., RUSCH, D., PETERSON, M., ROGERS, B. T. & Association Spec.Vol. Ser. 55, Chapman & Hall, London. KAUFMAN, T. C. (1996): Origin of the Arthropod mandible. KUKALOVÁ-PECK, J. (1985): Ephemeroid wing venation based Nature 380: 395. upon new gigantic mayflies and basic mor• SCHOLTZ, G., MITTMANN, B. & GERBERDING, M. (1998): The phology, phylogeny, and metamorphosis of pterygote in• pattern of distal-less expression in the mouthparts of crus• sects (Insecta, Ephemerida). Can. J. Zool. 63: 933-955. taceans, myriapods and insects: new evidence for a KUKALOVÁ-PECK, J. (1991): Fossil History and the Evolution gnathobasic mandible and the common origin of Mandibu• of Hexapod Structures. Pp. 141-179 in: CSIRO (ed.) The lata. Int. J. Dev. Biol. 42: 801-810. Insects of Australia, Melbourne University Press, Mel• SCHÖNMANN, H. (1979): Die Nahrungsaufnahme der Larven bourne. von Siphlonurus aestivalis Eaton. Pp. 293-298 in: PASTER• KUKALOVÁ-PECK, J. (1998): Arthropod phylogeny and 'basal' NAK, K. & SOWA, R. (eds.) Proc. 2nd Int. Conf. Ephemero• morphological structures. Pp. 249-268 in: Fortey, R. A. & ptera 1975, Panstwowe Wydawnictwo Naukowe, Wars- Thomas, R. H. (eds.) Arthropod Relationships, Systematics zawa, Krakow. Association Spec.Vol. Ser. 55, Chapman & Hall, London. SCHÖNMANN, H. (1981): Zur Kopfmorphologie der Ephemeri- LAUTERBACH, K. E. (1972a): Über die sogenannte Ganzbein- denlarven Siphlonurus aestivalis Eaton und Lepeorus goyi Mandibel der Tracheata, insbesondere der Myriapoda. goyi Peters. Zoologica 131: 1-51. Zool. Anz. 188: 145-154. SHAROV, A. G. (1966): Basic Arthropodan Stock with Special LAUTERBACH, K. E. (1972b): Die morphologischen Grundla- Reference to Insects, 271 pp., Pergamon Press, Oxford. gen für die Entstehung der Entognathie bei den apterygoten SHORT, J. R. T. (1955): The morphology of the head of Insekten in phylogenetischer Sicht. Zool. Beitr. N. F. 18: Aeshna cyanea (Miiller) (Odonata: Anisoptera). Trans. R. 25-69. Entomol. Soc. Lond. 106: 197-211. 178 A. H. STANICZEK

SMITH, E. L. (1988): Morphology and evolution of the hexa• STURM, H. (1987a): Das Paarungsverhalten von Thermobia pod head. Proc. XVIII Int. Congr. Entomol., Vancouver, domestica (Packard) (Lepismatidae, Zygentoma, Insecta). B.C.: 81. Braunschweiger naturkundliche Schriften, Braunschweig SNODGRASS, R. E. (1928): Morphology and evolution of the 2: 693-711. insect head and its appendages. Smiths. Misc. Coll. 81: STURM, H. (1987b): Zur Evolution des Paarungsverhaltens bei 1-158. den Machiloidea (Archaeognatha, Insecta). Mitteilungen SNODGRASS, R. E. (1935): Principles of insect morphology. der Deutschen Gesellschaft fur allgemeine und angewandte McGraw-Hill, New York & London. 667p. Entomologie 5: 104-107. SNODGRASS, R. E. (1938): Evolution of the Annelida, STURM, H. (1997): The mating behaviour of Tricholepidion Onchyophora, and Arthropoda. Smiths. Misc. Coll. 97: gertschi Wygod, 1961 (Lepidotrichidae, Zygentoma) and 1-159. its comparison with the behaviour of other . Pe- SNODGRASS, R. E. (1950): Comparative studies on the jaws of dobiologia, Jena 41: 44-49. mandibulate arthropods. Smiths. Misc. Coll. 116: 1-85. TUXEN, S. L. (1959): The phylogenetic significance of entog- SNODGRASS, R. E. (1951): Comparative studies on the head of nathy in entognatheous apterygotes. Smiths. Misc. Coll. mandibulate arthropods. 118 pp., Comstock Publishing 137: 379-416. Company, Ithaca, NY. TUXEN, S. L. (1970): The Systematic Position of Entog• SNODGRASS, R. E. (1952): A textbook of arthropod anatomy, natheous Apterygotes. An. Esc. Nac. Cienc. Biol. Mex. 17: pp. 1-363. Comstock, Cornell, Ithaca, NY. 65-79. SNODGRASS, R. E. (1955): The Dragonfly Larva. Smiths. VON KÉLER, S. (1963): Entomologisches Wörterbuch. Misc.Coll. 123: 1-38. xiii+774 pp., pis. i-xxxiii, Akademie Verlag, Berlin. SNODGRASS, R. E. (1960): Facts and theories concerning the WÄGELE, J. W. (1993): Rejection of the "Uniramia" Hypo• insect head. Smiths. Misc. Coll. 142: 1-60. thesis and Implications of the Mandibulata Concept. STANICZEK, A. H. (1996): The mandibular articulation of Zool. Jb. Syst. 120: 253-288. mayflies (Insecta: Ephemeroptera) and its implications for WALKER, E. M. (1931): On the clypeus and labium of primi• the phylogeny of the lower Pterygote lineages. Proc. XXth tive insects. Can. Entomol. 63: 75-81. Int. Congr. Entomol., Firenze: 2. WALKER, E. M. (1933): On the anatomy of Grylloblatta cam- STANICZEK, A. H. (1997): The morphology of Siphlaenigma podeiformis Walker. 2. Comparisons of head with those of janae Penniket (Ephemeroptera, Siphlaenigmatidae), and other orthopteroid insects. Ann. Entomol. Soc. Am. 26: its significance for the ground plan of the Baetoidea. Pp. 309-344. 536-549 in: LANDOLT, P. & SARTORI, M. (eds.) Ephemero• WALOSSEK, D. & MÜLLER, K. J. (1990): Upper Cambrian ptera & Plecoptera: Biology-Ecology-Systematics, Mau- stem-lineage crustaceans and their bearing upon the mono- ron+Tinguely & Lachat SA, Fribourg, Switzerland. phyletic origin of Crustacea and the position of Agnostus. STANICZEK, A. H. (2000): Der Larvenkopf von Oniscigaster Lethaia 23: 409^27. wakefieldi McLachlan, 1873 (Insecta: Ephemeroptera: WALOSSEK, D. & MÜLLER. K. J. (1998): Cambrian Orsten'- Oniscigastridae). Dissertation. University of Tübingen. type arthropods and the phylogeny of Crustacea. Pp. Germany. 139-153 in: FORTEY, R. A. & THOMAS, R. (eds.) Arthopod STRENGER, A. (1942): Funktionelle Analyse des Orthopteren- relationships, Systematics Association spec. Vol. Ser. 55, kopfes. Zool. Jb. Syst. 75: 1-72. Chapman & Hall, London. STRENGER, A. (1953): Zur Kopfmorphologie der Ephemeri- WHEELER, W. C. (1989): The systematics of insect ribosomal denlarven. Erster Teil: Ecdyonurus und Rhithrogena. Öst. DNA. Pp. 307-321 in: FERNHOLM, B. (ed.): The Hierarchy Zool. Z. 4: 191-228. of Life, Elsevier, Amsterdam. STRENGER, A. (1970): Zur Kopfmorphologie der Ephemeri- WYGODZINSKY, P. (1961): On a Surviving Representative of denlarven. Palingenia longicauda. Zoologica 117: 1-26. the Lepidotrichidae (Thysanura). Ann. Entomol. Soc. Am. STRENGER, A. (1973): Die Mandibelgestalt der Ephemeriden- 54: 621-627. larven als funktionsmorphologisches Problem. Verh. Dtsch. Zool. Ges. 66: 75-79. STRENGER, A. (1975): Zur Kopfmorphologie der Ephemeri- denlarven. Ephemera danica. Zoologica 123: 1-22. Author's address: Arnold H. STANICZEK, Universität Tübin• STRENGER, A. (1977): Zur Kopfmorphologie der Ephemeri- gen, Zoologisches Institut, Lehrstuhl für Spezielle Zoologie, denlarven. Proboscidiplocia skorai. Zoologica 127: 1-18. Auf der Morgenstelle 28, D - 72076 Tübingen, Germany. STRENGER, A. (1979): Die Ernährung der Ephemeropterenlar- ven als funktionsmorphologisches Problem. In: PASTER• NAK, K. & SOWA, R. (eds.) Proceedings of the Second In• Received: 10.03. 1999 ternational Conference on Ephemeroptera 1975, pp. Returned for revision: 13.07. 1999 299-306. Panstwowe Wydawnictwo Naukowe, Warszawa. Accepted: 02. 09. 1999 Krakow. Corresponding Editor: M. SCHMITT